SWUTC Research Project Description
Title of Project: Characterizing Information Propagation Through Inter-Vehicle Communication on a Simple Network of Two Parallel Roads
Project Number: 161005
Principal Investigator:
Bruce Wang
(979) 845-9901
P.I. Affiliation: Texas A&M University
Project Monitor:
Edward Fok
FHWA Resource Center
Transportation Technology Specialist
201 Mission St. #1700
San Francisco, CA 94105
(415) 744-0113
Project Status: Active
Date Started: 9/1/09
Estimation Completion Date: 8/31/10
Estimated Cost - Current Fiscal: $46,000
Estimated Cost - Total Planned: $46,000
Project Summary:
Project Abstract:
A new paradigm of self-organizing ad hoc mobile network on the highways (called VENet) has shown a significant potential in revolutionizing the network traffic operations and control. Enabled by the fast advancing computing and telecommunication technologies, equipped vehicles on an auto-net have wireless communication capabilities with a short transmission range. Auto-net essentially connects drivers on the road, captures transient traffic conditions, and enables drivers to make better informed decisions. It will improve the traffic system performance in terms of throughput, safety, resiliency, and integration.
The process of information propagation along traffic streams is fundamental to design of the VENet system. A good characterization of this process would enable adequate configuration of key system parameters such as transmission range, communication protocol, market penetration rate and roadside station spacing to assist the propagation. The objective of this project is to study the propagation process and analytically characterize it. We choose a special discrete network of two parallel roads for study in this project. We propose to develop an approximate equivalent process in order to develop formulas for the mean, variance of the propagation distance. A recursive formula for probability distribution of successful propagation will also be developed. Numerical test will be conducted to show the effect of transmission range and road separation on the information propagation. We hope this project will serve as a springboard to studying information propagation on a general urban grid network.
Project Objectives:
The goal of this project is to develop models and analytical results to characterize the stochastic process of information propagation as a function of such parameters as transmission range and spatial network traffic characteristics. The results will support development of a reliable and robust VANet. The result from this project will support research in several other areas such as development of protocols of communication, development of routing algorithms and advancement of parallel computing on ad hoc networks. In a larger context, the result here can also potentially support development of communication technologies among combat individuals in a battle field through short range wireless equipment.
The major objective of this proposal is to develop analytical models to characterize the process of information propagation in terms of its probability distribution, expected value and variance in the case of two parallel roads (a special form of traffic network). In particular, we propose to apply the Bernoulli process to develop approximate formulas for the above mentioned measures. The approximation is proposed due to the inherent complexity involved in a more accurate process.
Currently, we have already had some preliminary results of using the Bernoulli process to approximate the information process. Preliminary tests have shown high accuracy of some developed formulas.
Task Descriptions:
Task 1: Literature review
We will scan existing literature in several areas. 1. Literature on information propagation along one single road and that in two continuous dimensions (in the wireless communication area). 2. Relevant theories on Markov process and Bernoulli process. 3. Practices and theories in vehicle infrastructure integration in general. This scan will help find theories and methods that can be extended to the two parallel road case. Through the literature review, we will also be able to include more practical considerations in our research.
Task 2: Develop formulas for the propagation process along two roads
First, we will identify an equivalent underlying Bernoulli process. This Bernoulli process shall have the same key characteristics such as propagation distance in terms of expected value and variance.
With the identified Bernoulli process, which is easier to model mathematically, we will develop three formulas. One is the expected value of information propagation distance. The second is the variance of propagation distance. The last is the success probability of propagation distance. These three formulas will characterize the stochastic information propagation process, which are needed for research in several other research areas now. The formulas for expected value and variance will be approximate. The formula for success probability of propagation will be accurate.
In particular, we will use conditional expectation in probability theory to develop the expected distance of information propagation, and will use Eve’s rule to develop formula for the variance of information propagation distance. In addition, the probability formula will be developed in a recursive way.
Task 3: Apply the two-road formulas to one-road
Formulas for the one road case have been established earlier by Wang and his collaborators. In this task, we will show how the developed formulas for the two road case deal with the one road situation as a special case. Specially, we will provide an alternative proof to the formulas of variance in the one road case using our developed formulas in the two road case. This proof will illustrate inherent connection of analytical results between the two road and one road cases.
Task 4: Numerical test
The purpose of the test is to evaluate accuracy of the developed formulas for the expected value and variance of propagation distance.
A table of test scenarios will be generated combining various parameters of transmission range, road separation distance and vehicle density. The results from the developed formulas will be compared with ones from simulation in each instance. We will use results from large number of simulations in each instance as the benchmark.
In generating reasonable practical scenarios for test, we will define transmission range to be a basic unit of distance. All other measures of distance will be based on this unit. This is because distance is relative.
Task 5: Final report
Final research reports will give a complete description of the problem, approach, methodology, numerical test, findings, conclusions, and recommendations that are developed in the project.
Index Terms: