Transcript
Okay, good morning, good evening, good afternoon, and thank you so much for joining us today on this community live session. Optimize your wireless network for voice services on 9800.00:09Wlcsss. So I want to thank you all for being here. My name is Brooke Hammer. I'm the English community manager at community.cisco.com. We will have a chance to view this presentation, slides, and recording after the session within the community in a few days. I'll go over some more community things at the end of the.00:20Presentation. If you have any questions throughout your session, please do feel free to put it in the Q and A panel, and we will have question managers there to support you or you can wait until the end to ask questions in our Q and A portion.00:40Okay, today we have two experts that will be presenting. We have Jesus Hara, and Ernesto Floresce. Jesus is an ITSM from Mexico City campus alumni with a bachelor's degree on computational technologies engineering. He joined Cisco in 2017 as part of wireless tech in Mexico City where we.00:58Worked as team captain escalation engineer. Later he switched roles to business unit engineer where he has been working for the last four years. He has supported large events like Cisco Live, the Fifa world Cup 2022, and he was a speaker during Cisco Live 2024. Our Nesto Flores is the communications and electronics engineer graduated from the IPN.01:18With experience in international radio frequency regulations and technical provisions. Since 2022, he has been part of Cisco in the wireless tack area, specializing in the analysis diagnosis, and resolution of incidents in enterprise wireless networks. He has experienced as a technical speaker.01:38And wireless technology courses aimed at both internal and external engineers contributing to the standardization of best practices and we thank them so much for being here with us today. Today we'll have a couple Slido polls so you can either join using this QR code or I will launch them in the panel to your.01:57Right, as the session goes on. So again, feel free to join through this QR code on a different device or I will be launching them within the session. Your participation is greatly appreciated. We will also have a survey at the end of the session, so if you could please give us your feedback on the content, the speakers, and any ideas you have.02:17Having moving forward for another Cisco Live session, it is greatly appreciated.02:37Okay, give me 1 s. Let me launch our 1st one ok there we go, and it should be launching momentarily. Give me 1 s.02:44There we go. Okay, what voice and video applications do you have deployed at your job? Again, there's NO right answer for this. It is, you know, we just want to get a sense of what you guys are doing. So our options are zoom, Microsoft Teams, Webex teams, WhatsApp.03:07Discord or other. And right now we have a couple of folks coming in on that Microsoft Teams, Webex teams, and zoom. Oh, we got one other.03:22Yeah I'm gonna give a few more moments. We're still riding high on that Microsoft team and Webex teams. They're pretty close. Same really equal onto that Webex teams. Microsoft Teams. Oh, we got a few more for Zoom now.03:32And we'll let one more minute for another person is coming in with their votes here. Fantastic. So Webex Teams is taking the lead right now. We have Microsoft Teams and second. Zoom's coming in 3rd and other is 4th, and nothing for WhatsApp or discord at this.03:47This point so.04:07Okay over to you guys. Thank you so much for being here. Thank you Brooke. Everyone, this is nice meeting you and being with you on this morning, evening or afternoon for you. I have been working on the wireless business unit for the past couple years, but I have been working on wireless on Cisco for.04:12More than eight years now. So let me go over the agenda of what we're doing here. We're gonna be talking about an intermediate topics so there are some basic concepts that we will briefly just touch on while moving on forward. But the main purpose of this is to give you the best practices and our recommendations for your voice.04:32And video applications in your wireless network. So with that we're gonna be talking about 1st of quality of service as a theory on what it is, how it operates. Then we're gonna be moving to application visibility and control to know.04:52How are we monitoring how we are defining the application that runs through our network? After this, we're gonna be starting to move towards the configurations that are applicable to your improvement of your wireless voice application and video performance, which is 1st we're gonna be talking about standard.05:06Then we're gonna be talking about global voice over IP configurations that are available in your controller, what do they do? And then we're gonna be taking this global configuration and this roaming configuration and give you the best practices for your wireless LAN controller and your access point. After that, we're gonna be talking a little bit about how the call signaling works and how a call flows.05:25On your wireless network, all of these to end on a demo of how to decode a call using Wireshark, which is a very useful tool whenever you're trying to troubleshoot any issues with the voice quality or the video quality in your network. So.05:45Let's start 1st with a couple of terms for the glossaries. We are gonna be talking about some acronyms so it's better to have them handy if you're following this presentation. So, 1st WC for which stands for the wireless line controller.06:03AP that's our access point COS, which is the class of service, then the differentiated service code point, then EAP extensible if extensible authentication protocol. Qos which stands for quality of service, ABC, which is application visibility and control.06:20With the network based application recognition and then IP, which is the session initiation protocol.06:40So let's start with what is quality of service. This 1st couple of slides gonna be a little more theorical, but it's important that you understand the basis of and the foundation of what it is quality of services and how it impacts our networks.06:48So QoS. Qos is the way of how we prioritize traffic in our network. It's both applies both the wireless and wired networks. So QoS works by executing three actions to achieve this classification and prioritization. 1st, we're gonna be classifying the traffic.07:04Then we're gonna assign a tag to that traffic, and after we assign this tag, we're gonna queue this traffic on different queues depending on the priority to send one traffic or other cluster. So classification assigns the traffic to one of our four class labels that we have defined by the standards. It's voice, video.07:20Less effort or background.07:40After we classify our applications and our traffic, we use tagging and tagging it's canvas right depending on the standards that we use. The most common standard are COS and they assign these priorities that are later used by the QA mechanisms to send out the traffic over the.07:42The wire medium or the wireless interfaces. Now for a wireless network, we use the WMM for tagging, and this standard classifies the traffic in one of our categories that match the same ones as the normal passive class labels which are voice, video, best for our background.08:02A little bit more in depth on what are the tagging mechanisms that we have available. 1st class of service or CLS, it is based on the one P the one P standard which is a subset of the.1Q protocol, meaning this is going to be working on the layer two interface, the layer two protocol of the OC.08:23Protocol. Then this will take three bits in the header and this is what is going to be defined as a marking for our packets, and these markings will back to any of the values that we have for WMM or for QoS.08:43Now, differentiated service code point or works in the layer three, which means it's a part of the IP header on a six bit field that gives us a range of 64 values that we can use for tagging. However, we don't use normally all of these 64 values Rather we have come to a definition of some of.08:59The most common values like 46 which is for voice, 34 for video and zero for best effort.09:19Similarly, the ranges of the values mapped to a WMM or a QoS value. So it could be best effort, background, video or voice, so you have these tables at references, so you know whenever a packet is stacked, you know which category it matches on.09:25Why do we need QoS for our wireless networks? So in your normal wireless network, you can see your access point or your AP as a funnel. And this funnel will take all the traffic that your wireless clients are requesting and it's going to be using a 1st in 1st out type of input to scan.09:43Total the sending out those packets over the air.10:03So if you have clients that are browsing that are streaming that are sending messages that want to load a video or a webpage, all of those applications are gonna be queued on the same mechanism of 1st in 1st out, meaning some traffic that is more critical to latency can get deprioritized and sent.10:07Before some other traffic that can be offered or that can be asynchronously sent.10:27However, once we implement these QoS mechanisms, we are able to, the access point now classifies and defines the application based on the tagging that they receive from the wire network and will divide it on different queues and each queue will have a higher priority, you know that we're able to send more packets of the traffic that is more.10:32That needs higher priority and it's less more impacted to late from latency, but like our WhatsApp or our WhatsApp call voices or java voice calls or then our videos for Webex video conferencing.10:52As a secondary priority second priority and then the for some of the traffic that can be buffered that can take some more time to address or overcome this these latency mechanisms. So how does this look on our default QoS whenever we have a con.11:07Controller.11:27And we're gonna be talking about the a call from the wire phone to our wireless phone. And this is going to be an example of how the values and how our QoS works over a call. So in this scenario, we have the phone that is doing a call. So it's gonna start the.11:28The work phone is gonna start a call and it's gonna be sending out the packet. Here, the most important header for QoS is the DCCP because the phone is not aware of any tagging.11:48So we cannot define a cons of service value. So the phone is gonna be using this DCP value of 46, which stands for voice and then send it over to the wire network and then the switch that is connected to a controller will add this.11:59Tagging for a costs so it's gonna take the mapping of the 46 and map it to a cost value of six. Then this traffic is reaching the controller with the one Q header and the controller now has to encapsulate this packet into a cabo packet.12:15So whenever it does the header, the controller is going to add or copy the value from the CCP on the internal packet to the header. So we have this 46 value now again added on the header. This packet then travels all the way to the AP where the AP takes out this 46 value or the DCP header.12:30There and then whenever it puts it puts it over here.12:50It will do this mapping from the 46 to a double men value of six, which is the same value for voice. So that way the prioritization of voice is done whenever we are sending traffic over here. And keep in mind all this.12:54For our wired inter wired network, it's just my it's just seen as a header with a DCP value of 46, which is expedited forwarding or voice traffic. Now what happens whenever the phone, the wireless phone replies back? So the phone is gonna be sending its own packet with its own priority for WMM of six and.13:11VCP value 46. The access point now is gonna be taking this 46 value and mapping it to its own header. However, notice that the AP is now unaware of any internal mechanism to do the mapping for the 46.13:31To a.1Q value or tagging for QoS, so that value removes to zero now. And whenever it reaches the controller, then the controller is gonna untag the packet and it's gonna keep this 46 value of the CCCP to send it over to the phone, but the up value or the.1Q value for.13:48For COS will be set at zero. So that can cause some issues whenever we're traversing the network and this internal packet is left back into the wired interfaces.14:08So how are you gonna be overcome or how are we gonna manipulate those tagging and those labeling? We have on the 9803 different mechanisms to do so. 1st, we have the metal policies, which the metal policies are just predefined from the initial configuration of your controller. Those are already preloaded and as such, you cannot modify.14:20What they do.14:40What they will do essentially is just to allow some basic functionality to enable the AP to do higher GCP markings for and to do those mappings to all of their WMM and COS values. Then you have custom policies and custom policies are defined through class maps. So class maps are objects.14:41That will allow us to have greater granularity and control over which applications we want to do our remarking or which one to prioritize.15:01This can be done now per application manipulation. Additionally, we cannot only change the markings that we are gonna be using for the COS or the WMM value, but also we can do rate limiting and we can do all right packet dropping, so we can get rid of specific applications in our wireless networks.15:10And lastly, we have out of QoS, which out of QoS can be seen as a mix between the metal policies and the custom policies. It takes from the metal policies in the sense that are a group of class maps and policies that are predefined on the controller, so you don't need to do any manipulation.15:29And it takes from the custom policies that class map definition, so it works on specific applications. And the application that it's gonna be touching depends on the profile that we select because we have four different options. We have voice, video enterprise, and fast lane.15:46And one thing that you have to keep in mind is if you are going to use auto QoS, then you are not able to pile or to add an additional metal policy or custom policy on the profile where you added these auto QoS or you enabled out of QoS.16:02Then out of QoS becomes your only way to modify those mappings and those labelings. Now how where are we gonna be configuring each of these settings? If we're gonna take the metal policy route, the policy route gets applied directly on your policy profile. So you go to configuration.16:17Wireless attacks and profiles, then policy and if you select the policy profile that you want to add or modify, then you go to the QoS and ABC tab and you have your ingress and egress directions for your traffic and then there you will have a dropdown menu that has the four different options that you have. Platinum, gold, silver, and.16:36Which each one of these map to one of the values for QoS, background best effort, video or voice.16:57And this, again, it's just used to define this this maximum values that we're gonna be using to map it on the packets. Now talking about custom policies.17:06Custom policies are created by one or more class maps and they're apply on the same same QoS policy. This is configured on the configuration services QoS, and you will have the option to add a new policy profile not QoS policy profile.17:20This class map will have now a, it's constructed by 1st defining the application that we're gonna be manipulating and then the actions one that we're gonna be doing against that application.17:38The definition can be done per ABC, which we're gonna be talking a little bit about it later or it can be user defined. If you reselect the user defined, we have the option to match all traffic or any specific traffic and we're gonna be matching it depending on two things. It could be using the DSCCP.17:53Value of the original header of the packet.18:13Or if we know the IP addresses of the applications that are gonna be using or the, the servers that are gonna be using this application, we can define an access list with all of the list of IP addresses that we're gonna be modifying. Once we define the application that we're gonna be touching and that we're gonna be manipulating, we define the actions that we're gonna be taking about this.18:16Application traffic. We can do a remarking, so change the DSCP header type. We can drop the packet or this type of traffic or we can do a policy which is right limiting the kilobits per second that we are gonna be doing for this type of traffic.18:36Additionally, you have by default any time you create a new QoS profile, a default action or default class. This default class will have an action or will take an impact on all traffic that was not matched to any of our defined class maps.18:54So this is done, this is whether looked at if we do a graphic or a visual aid. So let's say we create our policy QoS policy map for a guest traffic, and we're gonna have two different actions. Like our 1st action is gonna be marking.19:13Packets from CCP we're gonna be reducing this value to 46 which is forwarding for voice to 34. Now it has less priority, it has a priority of video. Then we're gonna select all of the traffic that it's for YouTube or Netflix.19:29And when you're gonna drop this packet. Then if we select a different application, we're gonna rate limit that application to only 50 kbps. And then if the traffic did not match any of this, then we'll do a rate limiting of a hundred kilobits per second.19:45And after this, we go and apply this on our policy profile on the same setting as we apply our metal policies and we apply it either on the ingress or egress direction and for SSID or for the client. One thing that you have to keep in mind is that QoS policies work similar to an access list.20:01Meaning that the traffic is going to match only one classma. So if you have a very specific classma or specific traffic that you want to tag and identify.20:22That has to be defined 1st at the 1st class map and then have the more broad and the more gene general class maps definitions later. Otherwise, you're gonna be taking actions on traffic that you don't want to do.20:31For instance, now let's take a look, let's talk about auto QoS. So out of QoS, it's configured on the same menu that you configure your QoS policies, so service configuration services QoS. And in here instead of selecting a class map, you will own.20:46The only thing that you need to do is to select this auto QoS check or slider and after enabling this, you will be changed to this menu where you have to define the QoS macro, we have our again our four different options for enterprise voice case and then you will have a list of your available policy profiles.21:06Where you can enable this auto QoS profile on on them.21:26If you want to go and do it policy profile for policy profile, that can be done as well, but it it is more efficient to do it take this way. If you want to see which are the specific actions that each of these profiles or each of these macros does.21:32Then you can do so by running the show running configuration all command on your controller and that will have the definition of all these predefined profiles so you can define if that auto QoS profile it suits your needs for your business or for your use cases that you have deployed on your wireless.21:47Network.22:07So going back to group for the next question. Thank you so much. Let me go ahead and launch this. Okay, there we go. So as a network administrator, which mechanism have you implemented for visibility and control of traffic?22:11PC, NetFlow, wired QoS, wireless QoS or SDAVC.22:31So we have few answers coming in and so far ok, pretty even for ABC and netflow. We have a couple for wireless QoS and one for wireless wired QoS, sorry. None for SD ABC yet though.22:39Okay, and again it's staying pretty one net flow, two AVC, three wired, and four wireless, so we'll let another vote come in and yep, that's why we're sitting, and that's your answers for today.22:57Thank you, Brooke. So now I see that you are very familiar with the application visibility and control and which is good because we now want to we know how to manipulate the traffic, but we are missing how is this traffic defined or how the controller decides.23:14Which packet belongs to which application? That is when application visibility and control comes into play. Before going full into what IBC is, we need to talk about network based application recognition of MBAR. Because it's a proprietary Cisco mechanism that is used to identify applications that are flowing to our network.23:32Work.23:52And it works on layers four to seven, so we can detect traffic from different encrypted applications. So this is gives us greater visibility over the traffic and applications that are running to our network. There is one exception for encrypted applications if they're using point to point application and they are like SAS.23:53S AS application, then MBR is not able to identify it due to the encryption mechanisms and the algorithms that they use.24:13Now, we know that applications and definition of them are constantly changing. Every day we have, we see a new server, a new application, a new definition, a new port that is being reserved for a specific application. So how can we make sure that the controller stays up to date?24:24Then we have the ability for to be updated and to be aware of these new applications that are being deployed. We can do a protocol pack update which give us greater, more defined or more current.24:41Application definitions. If you're gonna be using an updated protocol pack on a redundant system, I have to make you aware that this protocol pack has to be installed on both controllers individually. Otherwise we're gonna be running into an issue where there is a synchronization.25:00Issue and the controller that and the standby controller will go into a reboot loop since it is not able to synchronize the the configuration from the.25:19Active controller. It's gonna be trying to constantly access a file that it does not have and that therefore it's going to start going into a reboot loop, so keep in mind that if you're gonna be updating your portable pack, you have to update it on both controllers separately.25:30The way that we can see which applications are being currently identified for our protocol pack is by running the command show IP mbar protocol ID, and this will give you a definition of all the protocols that are identified, which type of protocols they are running on and the ID that we're gonna be assigning.25:47In the case that we have modified the protocol pack version, we have three commands to determine the what we have configured and what is available for us to use. If you do show IPM protocol pack active, it will show you the currently running.26:07Protocol pack, and then if instead of running active, you run the loaded keyword, then it will give you a list of all of the protocol packs that are installed in your controller. And lastly, the indactive option is going to give you the protocol packs that are loaded in the controller, but they are not in use. So that way you can either.26:22Decide which command it's better for you to see like hey I can I want to see which one I'm running or if I'm not running one that might be updated.26:42But I it's not active yet. And how do we load these new protocol packs? So we have downloaded them from the software.cisco.com page, but where do we, how do we send it to the controller? The easiest way is to do it through the graphic interface. So if you go to configurations.26:51Services application visibility, you will see four headers is one is an option of at the action that you can take for the for application visibility. You select the upgrade protocol pack.27:11And once you select the protocol pack, you will see the list of applications that are currently recognizable and you will see the add button. Whenever you click the add button, you will be prompted with this as a window, and this window you can select the file that you have pre downloaded to your device and drag it here, set a destination which can.27:24With the boot flash or the hardest depending on the hardware that you're running.27:44And then just click upgrade and the controllers are going to automatically download the protocol back and going to be activating it. So now that we know what It's, we can now move on to talk about application visibility control or ABC.27:48Because ABC uses the foundations of and present traffic visualizations of our wireless client traffic of what our clients wireless clients are accessing. It allows us to create that as well the rules to limit the bandwidth and usage for applications that we have from the QoS policies because we now can.28:06Use the IBC definitions.28:26To manipulate the traffic in our network. ABC is a tool that is supported on all the AP modes, so it can be running APs that are in local flexconnect or fabric mode. With the difference, the difference being, where do we run the? If we have our access points in local mode, then the controller.28:28There is a sole responsible for running the MBR engine and doing that identification and classification of traffic.28:48However, if we're going to be using flexconnect most APs or fabric most APs, then the engine is run at the AP. So the AP is in charge of doing the analysis of the packet and then it's going to just report the con to the controller, the results of the traffic analysis.28:54And then the controller is going to present those visualizations or export that that information to a server a different server. If you're gonna be using QoS configurations with mbar, then keep in mind that these QoS configurations can only be done to physical ports, meaning that svis loopbacks are not applicable for you.29:11Need to use this ABC mechanisms.29:31And also keep in mind that ABC it's an engine it's a different engine from, so ABC will be mapped to the iOS XE protocol IOSXC version that you're running, which is different from the mbar, which can be independently updated to have more up to dated application definitions.29:35So in order to configure and enable ABC on your policy profiles for your wireless traffic identification, you have two options for configuration. 1st, you can go into configuration services application visibility and under the enable ABC tab.29:55You will see a list of all of the available policy profiles that you have defined on your controller, and then you can select which ones you want to have ABC enabled on. Once you select the click the once you click the blue arrow on the next to the policy profile name.30:11You will be added to the visibility tab, they will be added to the visibility and enabled tab, and then you have to define where are you gonna be presenting the information. Is it gonna be locally on the controller or are you gonna be sending this information externally to a server that collects this ABC information?30:26If you do it so if you decide to do an external collector, what's going to happen is that it's gonna be you're gonna be asked for an IP address, and then the controller is going to take this IP address and generate some configuration that is called a flow monitor and a flow exporter.30:43And this monitor and exporter will be unique for each policy where you add this collector. So now if you want to enable AVC for more than one policy profile, and these policy profiles are gonna be shared sharing the same collector.31:00You're gonna be having an increased number of exporters with repeated information. So let's say I want to enable the IBC on all these 140 policy profiles that I have, and I'm gonna be using an external collector. I will be having a hundred and 40.31:20Flow monitors now defined on my on my show running configuration which will make the file larger, even if we are repeating some of the collector addresses that I want to send the information to. So in order to avoid this issue, what we do now it's we can go into the.31:37The last header which is the flow monitors and define our custom exporters and monitors. If we 1st define the exporter on the 1st tab on this menu and we whenever we add an exporter, we just have to define the name that we're gonna be using, any description that is optional, and then if we.31:56Are you gonna be doing an external collector?32:16We just have to define the IP address and the port that is gonna be used for destination of the information for ABC. Once we have the exported defined, what we have now, the availability or to do it so it's to add those into a monitor, and the monitor.32:18It's just a group of different exporters that we want to send the information to. So we have a list of our exporters that we previously defined, we just add them to the monitor and then apply to the device. Then this master is ready to be used and mapped onto the different policy profiles that we have in our network and then we want to have ABC enabled on.32:38And lastly, talking about the last menu on our application visibility tab is to it's a defined policy. This is how where we can modify the different the default configurations for our applications. So applications are going to be defined or device.32:59Into three different categories, business relevant, business irrelevant or default. And this is going to have the different priorities or the different mechanisms of how we are going to.33:19The use them in our deployment. So by default, the controller is going to have different categories of applications and then we can go in there and do a definition or selection on where which type of classification we want to have for each one of these applications.33:29So we want to say some applications are not going to be used in our deployment, we can move them to the business irrelevant or the fault which it's also defines the type of action that we can take on the, on them. Additionally in this menu is where we can add custom applications in case we have our own internal.33:48How they develop.34:08Applications that are flowing through our wireless deployment and that are not part of the default ABC definitions, we can add them here. So the controller is now aware of these other custom applications. So now that we have defined which applications are business relevant, business irrelevant, and we have defined our monitors.34:09If we go back navigate back to the policy profile under the QS and ABC tab to the right side of the screen, then.34:29We're gonna have the flow monitor IPv four and IPV six and here's where we add the monitors that we define which will contain the exporters or where, where we are sending the information. And here is where we have one tag team. I noticed that with that on the poll, most of you are not aware or did not have not implemented as the ABC.34:38However, if you have access to a cattle center SDABC is a very powerful tool that will allow you to centralize aggregated analysis from multiple devices and sources. So you're not only looking at ABC information coming in from the controller, but you're also mapping it with information that this market protects from your switches and from your.34:58Servers and it consolidates all of this information into one single visualization point and that will give you better more information.35:18For your decision making. So you can now find do a finer tuning of your QoS policies knowing that there are more applications that you have to be aware of whenever you hand want to define the rate limiting or the marking of those applications.35:28Okay. And additionally, as the ABC will allow you to also take into consideration these custom applications in the way of how you visualize and you identify traffic over your network and your different devices. So what it's one alternative that we have to ABC, we.35:44Well, we can have ABC and that's a good tool. We also have NetFlow, and I noticed that you also have NetFlow implemented on your deployments, so that's a good thing, so let's talk about what is NetFlow. Netflow, unlike ABC, it's going to present visualizations, but the visualizations are gonna be created from what is known as an IP flow.36:04And an IP flow is going to be used to report in traffic network and analyze the trends that it goes over the usage of this network. The flows are created not only by the application definition, but also different inputs like the source and destination IP address of the packets, the protocols that the packets are using. They.36:24COS value or value along so many other factors that are part of the headers that we do in order to have a richer analysis of the hope of the packet or the traffic as a whole flow.36:44These flows are later exported to a collector, and this collector will take all of the all of these flows, aggregate them and process them to a visualization. So you all have a dashboard that has this visualization and this information about your traffic on your network.36:56Now, when we are talking about wireless networks, one thing that is very useful is that we enrich the flow of adding some information like the client MAC address in case your client uses a randomized Mac addresses we can identify these different randomized Mac addresses and include them on the.37:12As well as APs since our wireless clients are constantly moving across our deployment wireless deployment, we can see which APs they touched, which authentication mechanisms they use if they roam using a fast roaming mechanism or not, along with much more information that will give you a deeper analysis and more.37:31KPIs for you to analyze and do this, take this decision making.37:51And where do we configure netflow? It's configured under configuration services netflow, and we have your option to add a new NetFlow monitor and once you create a new NetFlow template, it will give you the option to have the collector address.37:55And then the exported, where this exporter with this collector is going to take the information, the source IP address where we're gonna be sending this flows from and also the sampling method which can be all the traffic that these that all the clients do or we can do other terministic sampling, so meaning we only send one every.38:15Other packet that we can define how many packets we want to analyze. Taking into consideration that doing this analysis and disaggregation information is going to take some CPU cycles to send to send it, to create the flow and send it to over to the exporter.38:35And lastly, we select the interfaces where we are gonna be monitoring. The advantage that network has over some of these MBR definitions and they can be applied to the whole SVI interfaces. So any interface that it's defined on the controller, it can be physical or logical.38:48Can be added as a way to monitor through NetFlow in the ingress and egress directions.39:08Now that we have all of these contexts, I'm gonna give the word to Ernesto so he can explain a little bit more of the actual configurations that we have available and how to make use of them. Thank you, Jesus.39:15I just want to thank you all for being part of this webinar. So Asusa was saying, we are going to start about to talk about fast roaming in our environments. In this case fast roaming, it is important because the quality of service in our.39:34Environments and application visibility and control that Jesus was mentioning for that are monitoring tools.39:54It is very important and that the connection between the endpoints and the access points should be reliable. In this case, faster and roaming through their environments needs to be determined by the protocol that should be implemented in our sites.39:59So we are going to pass to the next slide. Okay, just give me a minute. I'm getting some issues with this.40:19Could you please help me with that? Okay, so fast roaming it is the rapid transition between basic service sets that introduce a new concept in which the initial signal change with a.40:39New access point is performed even before the client moves to the destination access point. That means that there are a conversation between the endpoint when the client connects 1st to the 1st access point.40:59If the client wants to move to a different destination access point, there's there are an internal conversation before the client performs roaming, so this will save time in instead of start a new whole conversation with the destination access point.41:12The initial link protocol allows the client and access points to perform the transit key calculation in advance. Those keys are part of the exchange of the conversation before the roaming and these PTs are applied to the client and access points after the client responds to our reassociation request or it.41:29Responds to exchange with a new destination access point to send a new reassociation request with the destination IP.41:48So we have here some examples of the most useful wlans incorporated in your environments. In this case we have open SID, PSK and.one XSID. For the fast transition to make it work, we have different management frames and we have different packets that need.42:00Needs to be exchanged between access point and the destination access point within the client. In case of an open SSID, you can use four management frames to connect and perform that conversation. If you want to introduce authentication information.42:20Or when to introduce an encryption key. In this case it will be contained four management frames plus or equal frames to complete the four way handshake to to have eight frames in total. In case of.1X or if you are going to implement radius authentication for your SSID.42:36You need to make sure that choose the ORPTLS authentication for these clients and these will have the eight frames to for this authentication to be done.42:56On the following diagram, we are going to show you how this works without a fast transition. So we have a diagram into the left section that is a client show that will show a client roaming from access .1 to access .2. In this case, you will see that access .1 and access .2.43:13Has two circles that are meaning the coverage area for this access point and there are NO a quite a little bit of of overlapping between them that it is that the client will use to perform their roaming. In this case.43:34When client moves to AP one for the destination access point that should be AP two, the conversation needs to start from zero. E.g., you can see that for the 1st packet we can see the authentication request, response, resociation request, FT response, and on to complete the four way.43:49If client decides to move to a different AP, the conversation should start again. This could cause delay in our environments and in case that we are on a meeting like this webinar something that and you perform roaming without roaming, without fast transition enabled, you could face a.44:09Instruction in your service or any data packet lost.44:28So in comparison now with roaming with fast transition, we are able to see the same diagram that the client will run from AP one to AP two, but the difference is that there is a pre conversation between the endpoint and the destination access point in this pre conversation.44:36The client will send a reassociation request to the destination access point to perform the the exchange of the keys before the client runs to the destination access point. In this case it will save time and will will will give us a better transition to the destination.44:54Nation access points in our environment.45:14How we can configure fast transition in our environment. Fast transition is the IEEE standard for fast roaming and is compatible with most common mobile devices, but it is important that you need to take in consideration on your environments, the data sheet and the capabilities of your endpoints that you are going to connect to your wireless network beca.45:21Because if if there are any device that is not transition capable, it could cannot connect or you maybe face issues with roaming on your environment. So it is always a recommendation to verify the data sheet capabilities of your endpoints.45:41How how we can configure this into the controller? Well, the configuration is per WLAN and you will see it into the layer two section. There's a box that is called fast transition and you will see three different.45:56Options that you can choose to select the fast transition. We have adaptive enabled and disabled mode. In this case, adaptive mode is like when the controller dynamically enables or disabled fast transition.46:12Based on the client's capabilities. It is useful when you have Apple and Samsung devices that have some issues to connect, but that that was the design for these type of devices. So their recommendation should be now to use the enabled mode that in this mode it forces that just the client.46:32Things that perform fast transition can connect and if you do not want that any client or you have some legacy devices that cannot perform fast transition, you can choose the the to disable this feature. Also, there is another way to perform into the enabled mode. You can enable the fast transition.46:51And set the authentication key management for the fast transition and non fast transition capable clients that is called mixed mode and it is into the enabled section for the fast transition, this could guarantee that the clients that cannot negotiate fast roaming could connect and the clients that can do it will will.47:11Do it in your environment.47:31Now that we have a quick view of how fast transition works, how the monitoring and flow monitor in our environment needs to be checked for for for our roaming and for our monitoring in our environments, we are going to take a look into the global OIP configuration.47:36Operations that we can set into the controller.47:53To perform quality of service depends on what you want to accomplish. The 1st of all is the call admission control. The call admission control is a concept that applies only to voice traffic and not data traffic.47:57And not data. The implementation of requires that the client to send the traffic specification that is the packet to reserve bandwidth in your environments. Then we have the media parameters that are the the those are the multi.48:18Media parameters in the hundred that includes several computable settings to improve voice, video, and multimedia traffic over the wireless network. So in this section it is not just only focused on voice.48:38You can be focused on video and multimedia traffic or or the the quality of service that you want to accomplish on your media. For the media stream, the media stream is a function that makes the transit the the transmission of an IP multicast stream reliable over the year.48:49By converting the multicast frame into a unicast frame to perform a reliable connection to the clients, each media stream clients confirms that the reception of an IP multicast video stream or not and will say why you cannot receive that.49:08And finally, the EDCA profiles, the EDCA profiles is the access configuration that assigns preferential access to voice video and other types of quality of service on your environments. This can be modified per band on your on your nine 800 controller depends of the frequency that you want to prefer.49:27In your in your sites. For the call admission control, the configuration needs to be done. It can it can done globally or into the policy profile. In this example we are going to apply it by policy profile, and the section is at the end you can see that it's.49:46And you can enable call snoopings, send disassociate and send 40 a.50:06What Cap does is like regulate the voice quality by limitating the numbers of calls that can be active at the same time in the in one particular link. So it allows to regulate the bandwidth to consume by active calls on the link, but this is important that you need to take in consideration that this feature.50:12Could cause issues if you do not have enough bandwidth to provide to your clients because the SIP voice call a snooping section.50:32That it is the tick and you have the send disassociate and send 4086 vc. It it will send messages to the client that cannot connect so this configuration is used to track the bandwidth used for the voice calls per radio, but if after the maximum bandwidth is reached, the the bandwidth that you configured, new calls may not be.50:41Accepted into this radio and will send disassociate to new users or send the VC message to the to the phones. So you need to take care of that you need to take in consideration that.51:01For the media stream, you need to enable it by configuration wireless and there's the section that it is media stream. You need to enable the multicast direct enabled, and when you have this feature configured, you can send sessions announcements states. Those session announcement state is our messages, e.g., you.51:18URL email or phone, even a note. If someone cannot connect and and the client is not being able to access to the environment, it will announce an URL email or a message that you have pre configured on your controller.51:38And say maybe, ok, you cannot connect contact IT support or you cannot connect because the network is busy. So the client will know why they can do to in order to connect their their endpoints. It is important to mention that for media stream.51:57We must verify that the wireless controller has multicast service configured and that your network design can support multicast traffic because the net, the the controller will send that traffic over your network and your network needs to be prepared for to to forward that type of traffic on your environment.52:16And finally, we go to the EDCA profiles that these type of profiles as I said before depends of band that you want to accomplish and you can you can enable different types of EDCA profiles. In this case, this EDCA profiles parameters are into the section of the configuration.52:36Radio configuration and parameters. You can enable this feature per band. In this case we are going to focus in five years band and we are going to apply the optimized voice profile, but you have other profiles like custom voice, optimized voice or different types of traffic depends of your requirements.52:55Now that we have like a global view of the configuration that we have into the controller, my colleague Jesus will introduce you to the wireless line controller and AP best practices for a voice call. Thanks. Thank you, so.53:17So let's go over these configurations and these best practices. And the way that we're gonna be going over this, it's we're going from the most broad and the most global configurations and then we're gonna be digging down and deeper into the more specific and more prioritized configurations for.53:33The smaller objects like are your policy profiles.53:53So what is like bigger than the area environment where we are doing our voice calls or we're placing our voice calls, then I think we have to start from the design recommendations about the general RF recommendation and the RF environment that we want to have our clients perform.53:56Because if we don't have the proper media on the airspace.54:16For calls to have a better chance of doing a quality voice call, then we're starting already with some negative points into how our voice signaling will work. So in order to have an optimized environment, we need to ensure that the coverage area for.54:20Everywhere that we're expecting to have a call place has at least a two AP services with negative 67 dbmiss or stronger or signal. That way we can ensure that the voice quality or the connection for the voice packets it's gonna be optimal.54:40Then for the 5 ghz band since we are trying to use that as a less as a band that has less interference than 2.4 and it's more as wildly available for most of clients unlike the 6 ghz band.54:58Most of the client devices that use are or are they are designed for voice calls over IP have or are power aware of our environment. So we want to ensure that there's a minimum power level of at least eleven dbmiss for the 5 ghz band. That way we can ensure that our phones.55:13And our APs are talking to a similar power level. That way we don't have an asymmetric signaling where the phone might hear the AP louder than what the AP hears the phone at.55:33And if you have enough in your regulatory domain, you have enough channels, ensure that at least eight channels are available that will will give you the best channel optimized plan, a channel distribution plan.55:47For you to have the less co channel interference in your environment as we want to ensure that NO neighboring channels are next that are next to each other and that way we'll reduce this co channel interference.56:03There is a caveat here if you are in an environment where there are a high amount of EFS events like you are close to a weather station or an airport or a military base, then in that scenario, your.56:19Channel plan might be impacted by DFS events and we want to reduce the number of channel changes that can come into your environment. So we, you will want to avoid that you need to and you need to extend the channels if that is your scenario. Some more design recommendations, you have to ensure that.56:37It's NO higher than excuse me, than 1 %. The packet error rate is it's a the number of packets that are received with an error compared to the total number of packets that you have in your network. So ensure that the.56:57The errors are the errors can come from any physical challenge or from an AP that is very far away or any is configuration like multi pathing or fragmentation.57:17So ensure that the environment is as clean as possible from these obstacles whenever you are placing voice calls. The minimum data rate that we recommend for the mandatory configuration is per second. However, if possible.57:28You can have if you can have your network designed to at least have the 24 mbps as a I guess the expected data rate for most of your devices, that would be the ideal environment. However, this is I would understand that there's a construction of the number of density.57:46APs and clients that you will have.58:06However, if you have a very rock environment whenever there are the possibility for you to play access points, it's not as great or the IP distribution can take some distance between one and the other, then that scenario you can reduce to 6 mb per seconds to be enabled.58:08So that clients are able to connect and at least have this service for coverage.58:28For IP phones the majority will have one setting for the scanning mode and the scanning mode, what it will do is it will tell the phone to be constantly probing or probing at a recent frequent rate for any other APs and that way if we have continued scanning mode.58:34Most available, you can to enable it. This will allow more efficient roaming as the constant probing will tell.58:53Whenever an AP has a better signal for your phone to roam two. And that way we along with we can mark a more seamless and efficient roaming for your environment. The only compromise here is that continuous scanning mode will have a power.59:00Consumption so the battery life might be reduced if you are placing equal. However, if you.59:21Don't I know there's not the phone staying idle, then we can also provide you live by configuring what is called DTIM. So DTIM will allow the phones to stay on sleep mode for longer as the AP will not.59:27Required the phone to be awake or take it out of this idle state unless it has traffic that it needs to be sending out to the phone. So that way we can balance out the usage from whenever we have adaptive code place versus whenever the phone is on sleep mode to have it in that way.59:44Before longer with DTM configured.01:00:03And lastly, just a general recommendation, be aware of the software defect for known software on both your controller and devices. So it's updated to record any known defects. You have, you can take it on the gold star releases that we have on cisco.com. Currently for the ISXC controllers, we have.01:00:071715 four as the gold start release.01:00:27As well as for the phones usually the latest driver will have the most up to date for performance improvement. You can also like check out the different TAC recommended releases for any known defects or any major caveats.01:00:31Is that you need to be aware of for before implementing any new software on your deployment.01:00:51Now that we have these RF considerations taken care of and the design considerations taken care of, we can move now to talk about the configurations itself that we can manipulate in both our controllers. And now we also have to be aware of the switches towards our controllers connect to.01:00:57Because even if we do any of the work to do the identification of applications, the improvement of the queuing, the tagging of our packets, if the acquired network is not prepared to accept or honor the values that we have defined, then this information is going to be only done.01:01:17It's gonna be halfway or partially done. So, the thing that we now have to do is to ensure that the at least the wire ports that go to our access points and our controllers trust the backings that we are gonna be doing. And the configuration depends on the device that we are connecting to. So for the controller, since the con.01:01:37Controller uses VLAN tagging for the different clients traffic that we're gonna be manipulating for all your different we have to enable trust costs because we are again doing this marking on the layer two protocol, the.1Q VLAN header.01:01:57For the access points, we have two different configurations. If the access point is in local mode, then all of the traffic it's going to the controller, and as we saw in the 1st diagram, we are using instead of value from the internal packet and replicating into our header. So we want to prioritize this cap up.01:02:13Together and by that we need to enable trust the as our preferred mechanism for prioritization and tagging.01:02:32However, if the access point is in flex connect mode, then all of the traffic it's stays locally at the port switchboard where our AP is connected to, and our AP is now aware of any VLAN tagging and those header that it has to do. So in that sense, we have to do a trust cost because we are again using the vlan.01:02:41Heather as a way to have our tagging done. So that way our switch is aware and can do the proper queuing actions to prioritize our voice traffic.01:03:01Now we can move over to the controller configurations. And again, we're going from the global configurations all the way to the most specific configurations. So the 1st is to make a call out of control hack. As I mentioned, this mechanism, it's very aware of.01:03:12Bandwidth consumption and the bandwidth that's available. So.01:03:32In your environment, unless you have a very controlled set up where when where and when calls are placed, then the recommendation for best practice is to have disabled as a general marking.01:03:36Because we, it is hard to sometimes predict where the number of calls or the volume of calls will be done. So in that scenario, we can have a search of calls happening and some of these calls will then be blocked or will be receiving the network BC message on our phones as we, as a.01:03:52Cannot allocate any more resources or any of the expected bandwidth for all of these calls to be placed at the same time.01:04:12If you have a deep understanding of our network and the number of devices that are gonna be connecting per AP and the number of phone calls that are gonna be placed per APR per site, then when you can enable call admission control, but other than that the as a normal practice to just prioritize.01:04:18Guys.01:04:38Service or having this call come out or be completed, it is recommended to have it disabled. Now moving to the radio parameters with the here is we're gonna be modified the EDCA profile and we're gonna be having selected for optimized voice unless you're also having a.01:04:39The Usage of tele communication applications are video conference applications like Webex, Zoom, Microsoft Teams. In that case you.01:05:00Can use optimize for voice and video, but other than that this will allow us to have the best set of configurations and channel awareness for all of our voice traffic to be handled with proper priority.01:05:10Now moving to the RF profile, we're gonna be talking about the data rates about that we mentioned on the 1st slides for this section, and it's to have six six and 9 mhz as disabled and then 12 mb as mandatory.01:05:26Minimum mandatory data rate. This will allow us to reduce the logical cell for our phones and then our devices will be prevented from having from staying connected to one access point that might be far away.01:05:42And while not providing the best coverage or the best signal or the best data rates. So these will help us as well to do the roaming to a or moving to a more evidence to the access point instead. Keep in mind as well that this will require for you to do a side survey so you understand that there is.01:05:57Enough coverage for this minimum data rate to be available in the points where you are gonna be expecting to have calls placed.01:06:18Now we're moving forward to the more specific configurations. Now we're doing the double land profile. So whenever we create a new double land profile for our SSID that's specific for our voice deployment, we would want to have the 2.4 ghz span as disabled. That way we can save on some bigger.01:06:28In cycles for the access point as we are not advertising this network.01:06:48Unless you have some legacy devices that do not support 5 ghz, then it's when we will have it enabled for the greatest compatibility, but as of now, we want to ensure that the environment is as clean as possible so we are avoiding some of the most common interferes and non Wi Fi interferes.01:06:54On your network so having five years as the main provider for service will be beneficial the most beneficial. In terms of security, what we want to have is the mixed mode that spoke about on the earlier slides. We want to have a fast transition set as enabled.01:07:14With the one X AKM as enabled, we select.1X as the main, it's usually the most secure mechanism as it imply and doing some credential validation through AAA server.01:07:33And which some of the voice calls we need to have this protected, right? So besides enabling FT.1x, which is the necessary AKM or authentication key management or clients to be do performing fast transition, we want to have also.1x.01:07:50It's another one X chat to 56 enabled.01:08:10So we can allow some legacy devices that if they don't not support fast transition, they can still but they still need to place the voice calls. They can still make use of these features while maybe not having the best experience while roaming, but that way we still allow them to connect.01:08:13And within the advanced tab of the profile, we are going to call out two specific settings. One is the IE, so ironity permission element what it's going to be doing is for specific devices like Cisco IP phones, they will be able to provide more context and information about their.01:08:34Action of the signal and that way the controller is more it's better aware of the the presence of these devices and can adjust the channels taken into consideration information that it received from the phones. And then the other setting that we want to call out is off channel scanning deferred.01:08:54And afghanistanding refer is key in order to allow the access point to be aware of traffic that has priority and that cannot get be have the luxury of being interrupted. What of channel scanning defer does is it lets the AP know that.01:09:14It has to stay on the client serving channel if it receives packets with specific priorities. The priorities are going to be the ones that we define in this section. So we have the five and six priorities enabled. These five and six map to the voice WMM taggings.01:09:34And whenever the access point receive a5a packet with this tagging for five or six in the header, then instead of going to the off channel or to the next channel to do the scanning for any interferes, for any neighbor discovery.01:09:53Or any channel rogue environment or channel disc rogue discovery, it is going to prioritize handling this voice traffic that is present in our network and is going to be letting it go more fully and with the priority.01:10:11Because whenever the AP goes into one off channel, it will take some milliseconds to stop servicing clients which are key for us to avoid any latency or jitter in our voice communications. The time that we're gonna be deferring this off channel operation is defined by the scan deferred time, which by default it's a hundred milliseconds.01:10:32Which we can modify to increase it or decrease it to make the scanning differ a little more aggressive or less aggressive depending on the duration of our calls. But the hundred millisecond it's a good.01:10:52Medium of I would compromise on the off on the time that the AP spends off channel and the time that the AP needs to be aware of the system of these packets. Now within the policy profile, here's where we're going to apply most of our QoS and ABC.01:11:04Tagging. So 1st we're gonna be using the metal policy for platinum and platinum off for the direction respectively as a global SSID.01:11:24QoS policy. We do this because we want to again enable SDA as the DCDP markings for our voice traffic to be honored whenever we are doing the translation from the values to the DCP and COS packets on our head whenever.01:11:37If the AP sends the traffic back to the controller.01:11:56One tactic that we want to also call here is that if you are deploying your double lands and policy profiles with catalyst center, then you have to be aware that Catalyst Center creates a new policy profile for each site and each double land that it creates. It's important that you are aware.01:11:59Of this because if you have flow monitor enabled for ABC.01:12:18Then all of the and the policy profiles that are part of your policy tag will have to use the same flow monitor. This is a limitation that we have in our controller environment, so otherwise, then we're gonna be having a conflicting ABC configuration and then ABC will not be reported.01:12:23On your policy tax whenever you have this mismatch between two different monitors in your network. Additionally, one call out that we want to do here, it's for catalyst center provision business IDs.01:12:43If you enable QoS, Center is going to configure auto QoS instead of playing QoS SSID policies. So out of QoS has some, has had some hardening mechanisms that allows better performance and better awareness of the class maps applicant implementation and the rate limit.01:13:00Implementation that it does, in 1715 train codes for iOS xe. So if you are gonna be running out of QoS by yourself or through catalyst center, then the recommendation is to have 1715 on your controller as the minimum iOS XE software recommended.01:13:20Then under the advanced tab of the policy profile, we also want to ensure that our clients stay connected for the longest and with the less interruptions. We do that by 1st setting the session time out to a longer time than the default time.01:13:41So we usually would want to have the session timeout be as long as a normal shift for our wireless users is. So if you have clients that are gonna be staying on the network for 6 h, you can define the session timeout for these 6 h. That will allow the phone to stay connected for the longest and.01:13:56And any phone or any call that it's placed within, you will not have to have the phone reauthenticated every time that the session timeout completes. So during the ship, you only expect one of these authentications and then the next will be fast transition roamings.01:14:16From whenever the client moves. Then we also have to want to have client exclusion timeout disabled and this is to save time in case one of our phones is misconfigured and we address this quickly, we want to make sure that the phone is as quickly available to be back on.01:14:34Network and to be able to be placing calls or passing traffic.01:14:54So the client exclusion it's also very useful to have it disabled whenever we're doing any rollout of any new credentials or any new profiles for authentication. This way during this transition mode we are reducing the number of times that the clients that are yet to be provisioned.01:14:59Are excluded and therefore adding some burden into the configuration. So then as well we are gonna be using for the HTTP.01:15:17Sorry, then we're gonna be using the HTTP required. Do we have to have it disabled? This is because we also want to ensure.01:15:33That whenever the client is roaming, we take the least amount of time from whenever the client authenticated or passed this authentication process to be in the wrong state. Run state is the last status on the machine state for the client where we are able now to pass traffic and to let the client connect over the network or the wireless network. So what.01:15:48They are reducing or.01:16:08Avoiding the DHCP required, we avoid the con the controller for you to be aware or to expect to complete that follow way handshake with the DHCP server to grab any IP address which the phone might have an IP a static IP address or might have a reserved IP address that is going to continue for the whole.01:16:11Reminder of their session.01:16:31And last configuration that we want to call out here is for the policy proxy settings. This is configured at in the same advanced tab but closer to the bottom of that window, and we want to have our proxy enabled. This is key.01:16:34We are going to be reducing the time or the number of packets or overhead packets that have to go from our controller to the phone. So whenever an app request comes from a wire device asking for one of the clients that it's within the controller or within this SSID, then the controller is going to reply to this request.01:16:54Making the point of presence of that client, the switchboard going to the controller.01:17:14That way we are not forwarding these app requests through the access point, then to the phone waiting for the phone to reply, then the access point to send the reply to the controller and then for the controller to place that packet over the wired infrastructure. This were saving time processing CPU cycles.01:17:20And overhead traffic over there.01:17:40So that is the last one of our best practices that we have. So now we're gonna be moving to call signaling and flow, so we're gonna be talking a little bit about how a call is placed and how the packets exchange in one sample workflow or call flow that we have.01:17:44So in order to call about call signaling, we need to have to talk about session initiation protocol as this protocol is the one that's mostly used for the majority of the voice services. And this protocol what it's going to be doing, it's letting two devices or more devices in case we have.01:18:03One too many conversations or many too many conversations established as a session on a way where we, they can communicate over.01:18:24The protocol basically defines a series of requests and responses. And this response is going to have a code and this code is going to be pretty similar to HTTP, so if you have seen the HTTP four oh four or the 200 for an ok, then the similarly the phones are going to have different codes, based on the hundred.01:18:33It's two hundreds, three hundreds, four hundreds or five hundreds depending on the category of the response.01:18:53And then these are the it's a very robust protocol in the way that it defines a different number of scenarios and use cases through its own different requests and responses. The complexity of each call or each session is going to be dependent on the.01:19:00Number of devices that are part of the conversation and also if we have a centralized call manager or a centralized server that manages the.01:19:20Phones registration on the device registration on the lines. So we're gonna be for this example we're gonna be doing the most or the simplest call which is one phone talking directly to one other phone. There is NO call manager in between there we just have these two phones trying to.01:19:30Talk to each other.01:19:50So the 1st packet that we're gonna be sending is our phone A that's connected to an access point A it's gonna be trying to send a SIP invite to the phone B that it's on access point B So this invite is going to have the source and IP and destination of the.01:19:51A and source user and destination user or the phone, as well as some of the.01:20:11Media parameters like the session type, the actions that each that this phone can support each type of agent, which type of codecs for voice modulation they're gonna be able to use, as well as this any different call information that are specific to this session, including the session ID and their phone number.01:20:18Where that we are gonna be using.01:20:38Once phone B receives the, the packet, then its 1st going to send which is the ringing. And this is the tone that we hear on our phone whenever the receiver has not picked up the call. So it's the beef sound that place with.01:20:42While the phone is renewed on phone B Once phone B picks up the phone and answers the call, it's going to be sending a 200 ok response now, and this 200 ok it's going to look similar to the ZIP invite in the sense that it's going to have the codecs available, the session information.01:21:02The source user and destination user as well as the action that each user can support or this user can support and how much the session is going to call for or go for along with other different call details.01:21:20Once the phone A receives our ok, it just sends an acknowledge that it accepts the section on the session and then the, the call can be initiated as both of them understand each other and the parameters that they share.01:21:37We'll have a set of a stream of RTP packets that is our call that goes over both ways. And once we want to finish our call, then the one of the phones is going to send a zip by. The ZIP by is going to be in this case sent from the home B and it's.01:21:53Just going to be a request to finish and to close out the conversation.01:22:13Once the phone A sends a 200 ok, saying that it accepted the end of the call, then both phones are able to feed their resources to feed the bandwidth that is available and now both of them are.01:22:19Three or available to get a new call or to be placed any a different call now. So now that we have this context of how the call works and how the call is seen, Ernesto is going to take us from the last part of the, of this webinar where we are going to present.01:22:34The Demo how to.01:22:54The call the call is in Wireshark. Thank you Chris. Yes, ASUS was saying, it is important for us to understand the flow for the call signaling between the Cisco IP phones. In this case, we're going to use.01:22:56A feature that is at the code to the code and decrypt wireshark capture and we're going to take over the air packet capture on this to analyze the RTP streams in our environment to verify the quality of service of our voice callings.01:23:14In the in the next slide, we are going to see the diagram and and how the, the, the things can be done in defense of your encryption types. In case of an open, you don't require to perform nothing like more like the the overdue packet capture.01:23:33In case of a PSK SID, you need the overdair packet capture and the PSK password on plain text in order to decrypt the wireshark capture. In case of a.one XSID, it requires a different type of decryption method that requires a radioactive trace.01:23:52Of an endpoint that it is performing conversation and at the overdair packet capture. For further reference of this type of decryption, you can go further on the document that it is decrypt over the year packet capture in a.one access IDs. In our case, we are going to implement the PSK method, then this is the example that we are going to use. So, the.01:24:07Process of the real time transport protocol that is the RTP streams that my colleague has mentioned, we are going to use a Wireshark to analyze packet loss or the voice quality for our environment. This is useful when we are troubleshooting audio and video quality issues in our networks. For this, we are going to use another.01:24:27Controller 9178 access point, a call admission control that which is the cag that will allow the IP phones to obtain a SIP that will be the number of the ID that the access points will use to connect. And finally a PSK SSID with and without quality of service so.01:24:47The idea of this is that you will see the difference with and without quality of service applied into the environment.01:25:07So moving forward to the next slide, we are going to show you, so in our 1st example, we will use an SID that is called voice with a PSK encryption and NO quality of service configured. You can see here that the SSID is the call called voice tack and we have fast transition enabled as mixed mode.01:25:14And the password that we will use for the decryption of our capture will be Cisco 123 capital C.01:25:34Here we need to verify the policy profile because we are going to apply the quality of service in our policy profile. In this case you will see at the end of this command the show wireless policy detail for the voice that the auto QoS mode is select as known and into the QoS per SSID, there is not ingress or egress traffic.01:25:42Configured for quality of service. What we are going to do 1st is to explain you how to collect an over the air packet capture. As you can see, you have a network diagram that will show the sniffer access point of the AP service client or in case that you have a MacBook on.01:26:03Site you can also use a MacBook to perform.01:26:23This capture, it is important here to understand this tag tip that we are introducing here is that you need to collect the four way handshake between the endpoints and the access point in order to decrypt the capture because this is important as we mentioned before of the type of ssids that we want to accomplish in our environments, it depends on.01:26:25The of the management frames and the four way handshake to complete. So we want to collect that information because this that is useful to decrypt the capture.01:26:45So we are going to take the advantage of the MacBook in our environment. This MacBook or all the MacBooks has inside of it an embedded feature that is called wireless diagnostic tool. This wireless diagnostic tool will help us to collect and sniffer capture. The way to launch it is to go to the finder of your MacBook.01:26:58Search for wireless diagnostic tool, it will open a new menu and you need to go to the section that is called window, then click on.01:27:18Sniffer mode. With that it will open a new window that will ask for the channel and the, and the channel width that the client is using. In this case for our example, our client is using channel 36 with the bandwidth of 20 mhz.01:27:27Then once you have been selected this information, you need to start this packet capture and one is one you need to test and and once you have tested, you need to stop this capture and this capture will be in in the bar temp place for your MacBook.01:27:44What we need to do now is go to the Wireshark section and open the Wireshark software to decrypt our capture. In this capture you need to open the wireshark, go to edit preferences protocol. You need to select the wireless LAN protocol.01:28:05And enable the decryption box, then you need to select the decrypt keys, the decrypt keys, sorry. This section you need to put the PSK password and you must select the wPA-password and enter the password of your SSID in plain text. In our case, it's Cisco.01:28:22123. Now that we have selected our decryption method and we already said the decryption keys, we need to listen to the capture call and you need to click on the telephony section, then select RTP protocol.01:28:42And once you select this feature, it will open a new box that will show you the RTP streams that were capturing our over the air packet capture. Then you just need to select one of these ones and click on play stream in order to listen to your capture.01:29:02For our second example, we will use the same SSID that is voice, the same password is being enabled and the fast transition method as mixed mode is also enabled in our environment. The only difference is now the policy profile which is has now quality of service applied and you can see into the ingress and egress direction that we have platinum.01:29:21Up and platinum and the whoase mode is set to voice.01:29:41Now we are going to show you a demo of how decrypt this capture and how we can listen to our RTP streams in our environments. For the 1st part of the demo, we are going to analyze the the and we are going to analyze and decrypt the capture for the SSID without quite.01:29:48Service. 1st you need to verify that the equal key messages are being collected in your capture. We already verified that, so you just need to go now and select one of the packets, go to the pro protocol preferences, then selection wireless LAN and.01:30:09Verify that the decryption keys are already set. Then once it is set, you need to go to the telephony section RTP, RTP streams, and you need to select one of these ones. Now we are going to listen to the capture without quality of service.01:30:28Basically, we are performing a count from one to ten, but you can notice that it is.01:31:00It's not.01:31:07Clear and understandable the the the captures so even if you are able to hear the count from one to ten is not a good quality call at that moment. Now we are going to compare with the quality of service environment. In this case, we just need to verify again that the EPO keys are being set and collected.01:31:09Verify that the decryption key is already set again. Now go to the RTP streams and select one of the RTP streams after the quality of service being applied. Work? Two.01:31:30345-67-8910. Okay, so that that's the end of our demo and we'll.01:31:47We are now able to see and even more we are able to hear what is the difference and the comparison with.01:32:06And without quality of service enabled on our environment. This this this is an important tool that you can use in your deployments to verify that your quality of service and your voice calls, even you can use it for video, are good enough to.01:32:13Maintain a reliable connectivity. For the last part, we are going to verify the key takeaways that you need to to understand on your environment. Before implementation, you need to ensure the compatibility of end devices and applications. This is regarding two.01:32:31The data sheet of your endpoints, you need to make sure what type of endpoints are you going to connect to your wireless infrastructure in order to support all the capabilities. Also is important to mention that you need to consider the resources needed to ensure optimal distribution of the services. In case that you want to enable CAC, you need to take in consider.01:32:48The bandwidth that is the most important feature with that, with that service enabled.01:33:09Then for the best practices, we can configure the EDCA profiles to prioritize voice and video traffic per band. In our case, we are going to focus on 5 ghz band, but it depends of your requirements and you can tune EDCA profiles depends on the band that you want to to accomplish on site.01:33:15And enable fast transition to optimize the roaming, to perform seamless roaming on our environment. With this you will have a better connectivity for your clients and you will rely on the monitoring tools that my colleague Jesus said. Also apply the quality of service profile to desired service and depends on the.01:33:35That you want to accomplish in your environment in our case is voice. And after the configuration it is important to verify the the quality quality by decoding an air and over the air packet capture that that's the process that we saw.01:33:55With that you should be able to listen to the quality of service on your environment. What we want that gives you is the QoS key points that you need to accomplish in your environment.01:34:09You need to perform traffic classification and priority to maintain the desired service quality on your devices. You need to configure policies and parameters in governance dedicated to voice and video.01:34:24Depends on what you want to accomplish and always verify the compatibility with devices and applications to ensure effective roaming across your sites. And finally, but not least important continuous monitoring and adjustment to the tech and control network resources.01:34:39This means to have the engine and the service protocol up to date to rely on our monitoring tools.01:34:59With this information, we are moving forward to the Q and A section and it was a pleasure for us to give you all of this information. If you have any further questions, just please let us know. Thanks.01:35:09Wonderful. Thank you so much for a very informative session and for our two presenters for doing a wonderful job giving us all this content. So now I'd like to open it up for some Q and A So if anybody has any questions, go ahead and put in that chat window or that Q and A and we can.01:35:25Go over a few questions.01:35:44Additionally, we have our question managers Carlos and Jesus. I'm not sure if you would notice. Do you had any of the questions from the during the session? The only question I'm seeing is can we use Cisco Catalyst Center as an external exporter?01:35:52Yes, that is true that by defaulty center will set itself as an exporter for both ABC and netflow. So.01:36:13Either so yes, you can for sure use catalyst center or even prime infrastructure if you haven't migrated from prime infrastructure to catalyst center. Wonderful. I'm not seeing anything coming in. Thank you guys so much.01:36:27Just also real quick, if anybody is leaving the session, please do make sure that you fill out the survey at the end. It just helps us bring new sessions and new topics and everything to us. I'd also invite you to come over to community.cisco.com, where we invite you to connect, you can interact and collaborate. You can.01:36:47Join on the discussion boards for different conversations on these topics. If you have more questions, feel free to go to the community and post those questions where you will get peer to peer interaction, VIPs, and also internal Cisco responses. We actually just did a migration for the.01:37:07Community, so that will be over and we should come out of read only at the end of the day today. We will also have this recording and the slides and everything available to you by the end of the week in the community as well.01:37:27And I also encourage everyone to follow us on social. That is the best place to see where we have updated events and new topics and things that are coming to the community. We're on LinkedIn, X, Instagram, YouTube, Facebook, we're on blue Sky as well. It's not on here, but everything is under Cisco.01:37:43Community. So please feel free to join us on social. And once again, we thank you so much for your participation and joining us in our session today.01:38:02And that's about it unless there's any other questions that we have, but I don't see any of that came through. Okay, well wonderful. Thank you once again. Have a wonderful day and a great week. Thank you to our presenters. Thank you for our question managers, sorry, and we will see you on the next session. How.01:38:13Have a great day. Thank you everyone.01:38:32
About this Video
Author : Cisco Customer Success
Uploaded : May 28, 2026
AI summary
Shows how Cisco Catalyst 9800 Wireless Controller enhances voice and video over Wi-Fi with Quality of Service, Application Visibility and Control, NetFlow, and Fast Transition roaming. The recording covers EDCA profiles, Call Admission Control, Media Stream, and policy design, plus a live Wireshark demo decoding RTP to validate results. Includes recommended RF design, tagging, and policy profile settings to prioritize real time traffic.
- Provides step-by-step guidance to enable Auto QoS and Application Visibility and Control.
- Details configuring Fast Transition and EDCA profiles to optimize roaming.
- Demonstrates decrypting captures and assessing RTP quality with Wireshark.
Internal error
No related resources available.