Airforce Test Results for MeshDynamics
3-Radio Structured Mesh Module Network Design

The 3-Radio Structured Mesh Module was tested at a USAF laboratory under supervision by Air Force personnel and with approved network performance measuring equipment. The 3-Radio Structured Mesh module contained two 802.11a radios for backhaul and one 802.11b/g service radio.
 
Some of the tests conducted included:
 
Test 1: Bandwidth preservation over multiple hops
 
Despite much debate it has been conceded that the performance of single radio mesh networks degrades between 1/n (in the best case) and 1/2ⁿ (typical case) with each hop away from the Ethernet feed.
  
Meshdynamics uses a multi-radio backhaul technology that has minimal degradation with each hop.  To validate these claims, each Mesh node was isolated in such a way that it will only “see” the mesh node as indicated below. We then measured bandwidth available at each hop and also the bridge delay per hop.
 

 
Figure 1: Test Set up of 3-Radio modules to demonstrate bandwidth preservation.
 

 
Figure 2: Connectivity diagram of 3-Radio Backhaul radios.

Throughput is measured by using Chariot Test Software. The introduced delay is measured by using pings. The results are depicted in Table 1. Note: These are TCP/IP numbers, not raw data rate numbers.

Number of Hops from Root	Throughput in [Mbps]	Bridge Delay in [ms]
1	22-23 Mbps TCP/IP	Less than 1 ms
2	22-23 Mbps TCP/IP	Less than 2 ms
3	22-23 Mbps TCP/IP	Less than 3 ms

 
Table 1: WDS performance using 802.11a backhaul radios on 3 radio Structured Mesh Module
    
Figure 3: TCP/IP throughput measured at 1, 2 and 3 hops (3-Radio Structured Mesh).

Test 2: Bridge Delay over multiple hops
    
Figure 4: Bridge Latency measured at 3 hops is less than 3 ms. (3-Radio Structured Mesh).

Latency sensitive transmissions (e.g. VoIP and Video streams), are affected by the delay in wireless bridges. Recall that in the 3-radio Structured Mesh modules, there are dedicated service and backhaul radios. All data passes from the service radios to the backhaul radios via a high speed low latency wireless switch/router implemented in each Structured Mesh module. Over multiple hops, the latency introduced by the wireless bridge could be be significant. 

As shown in Figure 4, the bridge latency introduced - at the last hop - is less than 3 ms. Further testing indicated a bridge delay of  ~0.5 ms per hop for the 3-Radio Avila platform employed in these tests.
 
Test 3: Rapid Self-Healing of the Mesh
    
Figure 5: Mesh Topology was reconfigured in less than 2 seconds

A critical characteristics of high performance mesh networks is their ability to self configure rapidly in the event of node failure. One of the nodes was shut off - the system configured itself in less than 2 seconds.  Note: The delay is based on a sampling interval to decide if a node is indeed "down." It is adjustable.
     
Test 4: Dynamic control of mesh topology
 
The control layer is designed to automatically change the mesh topology when the signal strength changes between mesh nodes and connectivity performance is adversely affected. For details please see: Backhaul selection. To demonstrate the adaptive control over mesh topology a signal attenuator gradually changed the signal strength, thereby forcing the adaptive control layer to take action. The switch from the dotted connection to the solid line connection (Figure 7) occurred within one second of reaching the thresholds where bandwidth connectivity was adversely affected. 
   
Figure 6: Test Setup to see Mesh topology changing to support latency/throughput requirements.
   
Figure 7: How Mesh topology changed to support latency/throughput requirements.
 
Test 5: Interference management with multiple channels
 

One advantage of the multiple radio system is that each node operates on a different channel and interference is thus contained – since interference effects are restricted to one segment of the mesh.  This is a distinct advantage of the multiple radio system over single radio approaches.   This test demonstrates how the system reacts to interference on one channel- the mesh node switches to another backhaul parent, operating on another (interference free) channel, to manage connectivity performance in a proactive manner.

 
Figure 8: Switching channels to mitigate interference

An interferer was injected after node 2 in a 4 hop 2 radio Structured mesh network. The signal became so low that node 1 switched to node 3. The time will be measured between the time instance the SIR will reduce and the channel switching time from node 1 to node 3. Below the screen capture is depicted of this test. The test was done 3 times to see if it was reproducible. 
 
Summary 

Structured Mesh^TM provides distinctive advantages over traditional single radio approaches:

1) Minimal degradation of bandwidth over multiple hops.
2) Ability to modify mesh topology to meet dynamic performance requirements of latency/throughput
3) Ability to switch to other channels to mitigate interference effects in one part of the network. 

Voice over IP (VoIP) over Wi-Fi

See how bandwidth is distributed evenly over wide areas - with little degradation at each hop. Use VoIP phones to makes calls through the Structured mesh network to verify both latency and throughput performance. Latency is crucial for VoIP applications.

Compare the total cost of deployment of our approach vs. competing mesh products. Compare, for example, the number of Ethernet links needed to cover dense metropolitan areas for VoIP and data.

More Mesh Network Information:

• Various Approaches of Mesh Networking
• Hybrid Mesh Networks
• Wireless ISP Mesh Networks
• Technical Overview of Mesh Networks
• Advantages of Structured Mesh Network Modules
• Mesh Network Management Platform
• Benefits of a Mesh Control Layer
• Performance Analysis of Mesh Networks
• Test Results of Mesh Networks Real World Testing
• Deploying Wi-Fi Cellular Networks for Voice over IP Services