Skip to main content

Modeling and Analysis of Random Spatial Systems for 5G Networks

Periodic Reporting for period 1 - MARSS-5G (Modeling and Analysis of Random Spatial Systems for 5G Networks)

Reporting period: 2015-09-07 to 2017-09-06

With the adoption of 5G technologies and the increased reliance of intelligent transportation on communication, there is an increased need to ensure reliability of communication. The MARSS-5G research project focuses on the densification of nodes in next generation wireless communications. As networks become more dense, interference between transmissions becomes an important consideration. The MARSS-5G project considered as a case study vehicular communication and utilizes analytical tools to characterize the communication performance in a variety of operating conditions. An understanding of this performance is important to guide network design, which in turn has impact for traffic safety and efficiency. This project aims to develop several insightful analytical tools in order to: model, design, and analyze complex 5G networks. Tools from Stochastic Geometry and Planar Distributions for wireless networks have been utilized to enable the analytical derivation of these insightful performance metrics so as to assess the quality of service and assist in the design of complex networks. Such sophisticated metrics will be geared for applications in mobile and vehicular communications, thus eliminating the need for intractable, costly and time-consuming simulations. Through a novel approach to characterize communication reliability, the main conclusion of the work was the standard vehicular communication technologies such as IEEE 802.11p will be insufficient to meet the demands of future vehicular communication applications. This is because of the limited bandwidth and lack of interference coordination, which lead to low throughput and large packet loss in dense urban scenarios.
The work has led to three main results:
-a novel channel loss model for urban vehicular communication, suitable for tractable mathematical analysis
-a novel characterization of vehicular communication performance in urban and rural traffic, with finite deployment
-an evaluation of the fine-grained communication performance and contrasting with average communication performance

These results have been published in two conference papers at top-tier international conferences, one poster, one magazine paper, and submitted journal paper. These are listed in the publications.
The MARSS-5G project went beyond the state of the art through its novel tractable analysis, with a novel and realistic channel loss model, guidelines for network optimization, and an evaluation of the fine-grained reliability. The main impact of the work is that the average reliability of vehicular communication poorly represents the distribution of the actual reliability, especially when the communication is limited to a small area. This insight has profound implications for network design for high-reliability communication and will likely see increased adoption in the future.