As more vendors migrate to a mesh hierarchical design, and as larger networks are built using lightweight access points, we really need a standardized protocol that governs how lightweight access points communicate with WLAN systems. This is exactly the role filled by one of the Internet Engineering Task Force’s (IETF’s) latest draft specification, Lightweight Access Point Protocol (LWAPP).
With LWAPP, large multi-vendor wireless networks can be deployed with maximum capabilities and increased flexibility. Well…okay, this is mostly true. No one, and I do mean no one, has actually deployed a Cisco and Motorola network within the same company and is sitting back smugly saying, “Dude, this is really cool!” They’re saying something loud for sure, but it isn’t that! Cisco is Cisco and Motorola is well, not Cisco, and even though they supposedly run the same IETF protocols, they just don’t seem to see the standards exactly the same way. Basically, they don’t play well with each other.
So, let’s say we’re using only Cisco. (Hey, we already have an unlimited budget here, so why not put in all Cisco too, I mean, this is a “Cisco” book, right?)
Okay—so Cisco’s mesh networking infrastructure is decentralized and comparably inexpensive for all the nice things it provides because each node only needs to transmit as far as the next node. Nodes act as repeaters to transmit data from nearby nodes to peers that are too far away for a manageable cabled connection, resulting in a network that can span a really large distance, especially over rough or difficult terrain. Figure 1 shows a large meshed environment using Cisco 1520 APs to “umbrella” an area with wireless connectivity:
Plus, mesh networks also happen to be extremely reliable—since each node can potentially
be connected to several other nodes, if one of them drops out of the network because of hardware failure or something, its neighbors simply find another route. So you get extra capacity and fault tolerance by simply adding more nodes.
FIGURE 1 Typical Large meshed outdoor environment
Wireless mesh connections between AP nodes are formed with a radio, providing many possible paths from a single node to other nodes. Paths through the mesh network can change in response to traffic loads, radio conditions, or traffic prioritization.
Cisco LWAPP-enabled mesh access points are configured, monitored, and operated from and through any Cisco Wireless LAN Controller deployed in the Cisco Mesh Networking Solution—and they must go through a controller, which is why having redundant controllers is an absolute necessary.
Let’s define a couple terms used in mesh networks:
Root Access Points (RAPs) This access point is connected to the wired network and serves as the “root” or “gateway” to the wired network. RAPs have a wired connection back to a Cisco Wireless LAN Controller. They use the backhaul wireless interface to communicate with neighboring Mesh APs.
Mesh Access Points (MAPs) The Mesh APs are remote APs that are typically located on rooftops or towers and can connect up to 32 MAPs over a 5GHz backhaul. During bootup, an access point will try to become a RAP if it’s connected to the wired network. Conversely, if a RAP loses its wired network connection, it will attempt to become a MAP and will search for a RAP.
A typical mesh network would include the devices shown in Figure 2.
In Figure 2 , you can see that there’s one RAP connected to the infrastructure, and the MAPs connect to each other as well to the controller through the RAP.
But we’re not quite done with this yet—I want to explain one more mesh term before we get into wireless security: AWPP.
FIGURE 2 Typical devices found in a Cisco mesh network
AWPP
Each AP runs the Adaptive Wireless Path Protocol (AWPP)—a new protocol designed from the ground up by Cisco specifically for the wireless environment. This protocol allows RAPs to communicate with each other to determine the best path back to the wired network via the RAP. Once the optimal path is established, AWPP continues to run in the background to establish alternative routes back to the RAP just in case the topology changes or conditions cause the link strength to weaken.
This protocol takes into consideration things like interference and characteristics of the specific radio so that the mesh can be self-configuring and self-healing. AWPP actually has the ability to consider all relevant elements of the wireless environment so that the mesh network’s functionality isn’t disrupted and can provide consistent coverage.
This is pretty powerful considering how truly dynamic a wireless environment is. When there’s interference or if APs are added or removed, the Adaptive Wireless Path Protocol reconfigures the path back to the rooftop AP (RAP). Again, in response to the highly dynamic wireless environment, AWPP uses a “stickiness” factor to mitigate routes that ensure that an event, such as a large truck passing through the mesh causing a temporary disruption, doesn’t cause the mesh to change unnecessarily.
With LWAPP, large multi-vendor wireless networks can be deployed with maximum capabilities and increased flexibility. Well…okay, this is mostly true. No one, and I do mean no one, has actually deployed a Cisco and Motorola network within the same company and is sitting back smugly saying, “Dude, this is really cool!” They’re saying something loud for sure, but it isn’t that! Cisco is Cisco and Motorola is well, not Cisco, and even though they supposedly run the same IETF protocols, they just don’t seem to see the standards exactly the same way. Basically, they don’t play well with each other.
So, let’s say we’re using only Cisco. (Hey, we already have an unlimited budget here, so why not put in all Cisco too, I mean, this is a “Cisco” book, right?)
Okay—so Cisco’s mesh networking infrastructure is decentralized and comparably inexpensive for all the nice things it provides because each node only needs to transmit as far as the next node. Nodes act as repeaters to transmit data from nearby nodes to peers that are too far away for a manageable cabled connection, resulting in a network that can span a really large distance, especially over rough or difficult terrain. Figure 1 shows a large meshed environment using Cisco 1520 APs to “umbrella” an area with wireless connectivity:
Plus, mesh networks also happen to be extremely reliable—since each node can potentially
be connected to several other nodes, if one of them drops out of the network because of hardware failure or something, its neighbors simply find another route. So you get extra capacity and fault tolerance by simply adding more nodes.
FIGURE 1 Typical Large meshed outdoor environment
Mesh is a network topology in which devices are
connected with many redundant connections
between nodes.
connected with many redundant connections
between nodes.
Wireless mesh connections between AP nodes are formed with a radio, providing many possible paths from a single node to other nodes. Paths through the mesh network can change in response to traffic loads, radio conditions, or traffic prioritization.
Cisco LWAPP-enabled mesh access points are configured, monitored, and operated from and through any Cisco Wireless LAN Controller deployed in the Cisco Mesh Networking Solution—and they must go through a controller, which is why having redundant controllers is an absolute necessary.
Let’s define a couple terms used in mesh networks:
Root Access Points (RAPs) This access point is connected to the wired network and serves as the “root” or “gateway” to the wired network. RAPs have a wired connection back to a Cisco Wireless LAN Controller. They use the backhaul wireless interface to communicate with neighboring Mesh APs.
Mesh Access Points (MAPs) The Mesh APs are remote APs that are typically located on rooftops or towers and can connect up to 32 MAPs over a 5GHz backhaul. During bootup, an access point will try to become a RAP if it’s connected to the wired network. Conversely, if a RAP loses its wired network connection, it will attempt to become a MAP and will search for a RAP.
A typical mesh network would include the devices shown in Figure 2.
In Figure 2 , you can see that there’s one RAP connected to the infrastructure, and the MAPs connect to each other as well to the controller through the RAP.
But we’re not quite done with this yet—I want to explain one more mesh term before we get into wireless security: AWPP.
FIGURE 2 Typical devices found in a Cisco mesh network
AWPP
Each AP runs the Adaptive Wireless Path Protocol (AWPP)—a new protocol designed from the ground up by Cisco specifically for the wireless environment. This protocol allows RAPs to communicate with each other to determine the best path back to the wired network via the RAP. Once the optimal path is established, AWPP continues to run in the background to establish alternative routes back to the RAP just in case the topology changes or conditions cause the link strength to weaken.
This protocol takes into consideration things like interference and characteristics of the specific radio so that the mesh can be self-configuring and self-healing. AWPP actually has the ability to consider all relevant elements of the wireless environment so that the mesh network’s functionality isn’t disrupted and can provide consistent coverage.
This is pretty powerful considering how truly dynamic a wireless environment is. When there’s interference or if APs are added or removed, the Adaptive Wireless Path Protocol reconfigures the path back to the rooftop AP (RAP). Again, in response to the highly dynamic wireless environment, AWPP uses a “stickiness” factor to mitigate routes that ensure that an event, such as a large truck passing through the mesh causing a temporary disruption, doesn’t cause the mesh to change unnecessarily.
No comments:
Post a Comment