Distributed Mutual Exclusion Algorithms for Intersection Traffic Problems

Harisu Abdullahi Shehu,Rabie A Ramadan,Md Haidar Sharif

doi:10.1109/access.2020.3012573

Harisu Abdullahi Shehu, Rabie A Ramadan + Show 1 more

Open Access

https://doi.org/10.1109/access.2020.3012573

Copy DOI

Abstract

Conurbations around the globe are encountering the challenges of traffic congestion. Miscellaneous smart systems have been developed to help control and improve traffic flow in a cost-effective and measurable way. Yet, the existing systems solve the traffic congestion problem with cost-intensive traffic lights and mostly handle emergency cases poorly. In this article, to minimize the traffic congestion problems for any kind of vehicles without using traffic lights, we propose three deadlock-free algorithms namely: (i) Mutual exclusion algorithm based on single instruction (MEASIR), (ii) Mutual exclusion algorithm based on priority (MEAPRI), and (iii) Mutual exclusion algorithm based on multi-agent systems (MEAMAS). Communication inside a group is accomplished via a queue structure, while an external element (e.g., a router) is used for internal communications. Besides the depiction of experimental and simulation results, a complete statistical analysis has been performed to compare the performance of MEASIR, MEAPRI, and MEAMAS with their alternatives. Our proposed deadlock-free algorithms are not only efficient but also functional with a computational cost of $O(n)$ to enter the critical section, where $n$ represents the number of all vehicles in a particular intersection.

Highlights

Cities and towns around the globe are coming upon the challenges of traffic congestion and increasing emissions
To minimize the aforementioned existing problems at traffic intersections for any kind of vehicles flow, we propose three different mutual exclusion algorithms namely: (i) Mutual exclusion algorithm based on single instruction (MEASIR), (ii) Mutual exclusion algorithm based on priority (MEAPRI), and (iii) Mutual exclusion algorithm based on multi-agent systems (MEAMAS)
We have considered the rule of the mutual exclusion algorithm, which states that only one vehicle is allowed to be at the critical section (CS) at any given time

Summary

INTRODUCTION

Cities and towns around the globe are coming upon the challenges of traffic congestion and increasing emissions. Raymond’s lock-based algorithm [34] for mutual exclusion allows each node to have only one parent and maintains a FIFO (first-in-first-out) queue of request is applied to tree on distributed resources It is unknown if this same idea will function to solve traffic congestion problem. The existing problem incurred by the rule of mutual exclusion algorithm has been resolved completely in our proposed multi-agent systems (MAS) based MEAMAS, where more than one vehicle can pass the intersection at any given time. As we are using the identical idea of Raymond algorithm, the computational complexity of our algorithms tends to be lower than their alternative algorithms This could be one of the main reasons why our three comprehensively developed algorithms can achieve better performance for solving traffic congestion problem at the intersection. The rest of the paper is organized as: Section II briefs our proposed algorithms; Section III discusses related works; Section IV presents our system models; Section V illustrates our proposed algorithms; Section VI reports both experimental and simulation results followed by a detailed discussion including statistical tests as well as limitations of our algorithms that could be addressed in further study and Section VII concludes the paper

BRIEF IDEA OF OUR ALGORITHMS

OUR PROPOSED ALGORITHMS

2: Send token to router with higher priority among

RESULTS AND DISCUSSION

FUTURE WORKS

CONCLUSION