الفهرس 
Chapter 1Only 14 pages are availabe for public view
 |  | from | 99 |  |  |
| | | from | 99 | | |
Abstract
Road traffic congestion is one of the major problems facing both developed and developing countries alike. With the number of vehicles on roads exceeding billion vehicles and increasing every day, road traffic congestion is going to exaggerate. Road traffic congestion is one of the main reasons around the world for delay, time loss , fuel consumption, and frustration. We present our system StreetTel as a novel RF-based traffic detection and identification system. Compared to the current approaches for traffic estimation, StreetTel is low-cost, does not disrupt traffic during installation, works for non-laned traffic, and does not require active user participation. Our approach is based on the fact that the presence of an object in an RF environment affects the signal strength and hence can be used for detecting and identifying objects. We present the StreetTel system architecture and how it uses statistical techniques, based on the mean and variance of the received RF signal strength, to detect the presence of objects and differentiate between humans and cars to reduce the traffic estimation outliers, classify the level of road congestion, and estimate vehicles speed. Implementation of StreetTel on standard RF equipment shows its capability of detecting the presence of objects, identifying their type, classifying road status and estimating vehicle speed with high accuracy highlighting its promise for different vehicle-related applications. Our proposed system, StreetTel, showed through experimentation a good performance. Our system, StreetTel, is composed of two main subsystems, namely: MONITOR and TIMELESS. The MONITOR subsystem is an RF-based object detection and identification system. The MONITOR subsystem evaluation showed that we can achieve accurate detection with zero false negatives and positives concurrently. The system designer can tune the parameters to achieve different identification probability for humans and vehicles. The TIMELESS subsystem is an RF-based road traffic congestion estimation. TheTIMELESS subsystem showed a 90% probability of correct classification with congestion false positives equal to 0.113 and free flowing false positives equal to 0.079.