Methodology

The researchers conducted field studies and simulation tests to evaluate the reliability and effectiveness of two wireless technology systems. At the subsystem level, the researchers evaluated performance and reliability using field tests in highway environments. The factors tested included transmission power, modulation rates, highway terrain, and foliage obstructions. At the system level, the researchers conducted a simulation study to assess throughput per device according to different network topologies. A performance cost analysis was also conducted.

Field Tests

For the Wi-Fi field tests, the researchers collected data for communication between two adjacent nodes using iperf, a network testing tool, to measure throughput. The tests were conducted between August 2008 and December 2009 at two locations along South Carolina Highway 93. Different distances were tested, according to various modulation rates and transmission powers. Signal strength and signal-to-noise ratios were recorded and measured. Each test was 120 seconds, and throughput and signal strength measurements were recorded at each second.

For the WiMAX field tests, a study was conducted for fixed operation and nomadic operation.

For the WiMAX fixed operation test, seven locations were selected to measure the upstream and downstream throughput. Ten one-second samples were collected using the iperf network testing tool. For the WiMAX nomadic operation test, data samples were collected every 1 second to compare how the location of the base-station would affect the signal strength. A coverage measurement tool developed by Clemson University was used to assess the coverage of the WiMAX network.

A field test was also conducted from May to July 2009 using a traffic camera on Highway 39 to assess quality requirements of online traffic surveillance. Jitter and packet rates were mapped to assess video quality. The incoming video was recorded in 60 second intervals, measuring the frame rate and bandwidth. The arrival time of each data packet was used to calculate the jitter.

Simulation

A communication simulator was conducted to study the performance of wireless technologies under different topologies and environmental conditions. With ns-2, a network simulator was used to model the system and analyze the saturated throughput and the successful delivery ratio. Additionally, an integrated traffic simulation was conducted using PARAMICS. The integrated simulation assessed incident detection rate, false alarm rate and communication latency.

Performance Cost Analysis

A performance cost analysis was conducted based on the simulation results. The analysis was conducted using the Greenville, South Carolina network as an example. Cost effectiveness was computed as the throughput/cost ratio. For the purposes of the cost analysis, dedicated short range communications were also considered

Findings/Results

The field tests and simulations provided the following findings regarding parameters for higher performance:

• Wireless communication between traffic sensors significantly degrades after 300 feet (when using an off-the shelf radio with transmission power up to 70mW); this performance varies according to transmission power and modulation rates.
• Line-of-sight greatly affects connectively level for WiMax.
• Sensor spacing of up to .5 miles for Wi-Fi provided desirable results for incident detection.
• A jitter threshold of one second and a packet rate of 23 packets per second were sufficient for acceptable quality of surveillance video

In terms of performance-cost, the Wi-Fi mesh network was found to have the highest throughput-cost ratio (109 bits/dollar) for supporting traffic surveillance. However, WiMAX provides the greatest throughput (9.15Mbps per device).