Reliable Bandwidth Conservative Queue-End Detection and Warning System Using Smart Phone Collaboration Techniques

Abstract

The goal of this thesis is to ensure public safety. Two of the major components of public safety are: threat detection and notification. This thesis provides threat detection through sensor networks, collaboration of devices, and notification through three types of warning messages using cell broadcast technology. Most emergency conditions, be it natural or manmade, either directly or indirectly create congestion both on the road and in the telecommunication network. The more severe the road congestion is, the more vulnerable the tail end of the congestion is to rear end collisions. In order to avoid rear end collisions, it is important to know the end location of the traffic jam (i.e. the end of the queue) hence the name, queue-end (q-end) problem. If the exact q-end location is detected and made known to the drivers in a timely manner, many more accidents and incidents can be avoided. The q-end problem is resolved in three major steps: congestion detection, exact q-end location determination, and warning. Dynamic nature of the q-end problem makes it an iterative process. The design objective requires the traffic monitoring and warning system to be reliable, available, and use conservative resources. A middleware has been designed that will allow important messages to reach affected areas during network congestion. Cell broadcast is a one-to-many communication method, also known as Short Message Service-Cell Broadcast (SMS-CB). SMS-CB uses non traffic barring network resources and is unaffected by network congestion and most partial failures. This system also takes advantage of other wireless communication methods (i.e. Zigbee, Wifi, Bluetooth etc), internal stand alone resources, and collaboration capabilities built into smart phones. The proposed solution is designed as a multi-tiered cloud architecture system. The solution ensures reduced network traffic, multi-level backup mechanisms, as well as localized computing and control. The middleware supports both centralized and standalone operations, minimizing the impact of catastrophic network failure. The queue end or danger detection will be performed through a collaboration of sensors or user devices, which will send information to designated location for validation. In most situations, the validated information will be sent to the users via cell broadcast technology. Simulations and actual road tests have been performed to prove that smart phones are capable of detecting and sharing congestion with other smart devices and that they have the ability to receive and process warning messages to inform the end user of potential danger (i.e. location of queue-end). The middleware solution is suitable for day-to-day traffic congestion, accidents (i.e. q-end problem) when telecommunication systems are fully functional, as well as during any other major disasters (i.e. terrorist attack, major natural catastrophe, industrial accidents or other major issues) when telecommunication systems may be congested or malfunctioning.