NETWORK PLANNING: Deploying an Effective Backhaul Network
Network planning to support mobile communications is a complex, time consuming, and hence expensive process. The planning use case may involve, among others, any of the following possible scenarios: a cell phone provider deciding where to install the towers for the backhaul network, placement of access points (APs) for a vehicular or WiFi network, a military operation planning a network deployment to support mission command and control (C2), or a test range deploying range instrumentation for accurate Real Time Casualty Assessment (RTCA) in a military training or test event.
In each of the preceding cases, the goal of the network planner is to deploy the backhaul network such that it provides adequate coverage and bandwidth to deliver the end-to-end application traffic that originates from the mobile units, while optimizing use of the most expensive communication assets such as towers or APs for a mobile service provider, or airborne relays and/or communication gateways in a military deployment. We will use the term backhaul network to refer to the network connecting the towers, gateways, or access points.
NETWORK SIMULATIONS: Using Digital Twins to Understand Performance
Network simulations or network digital twins provide a partial solution to this problem by creating a digital simulation model of the proposed backhaul network, along with the operating environment and traffic load, and ultimately are able to use the digital twin to compute relevant performance metrics and cost. However, there remains a gap between assessing performance of a specific network deployment and generating an optimal network plan that meets such objectives subject to specified operational and cost constraints. The problem is further compounded if cyber and electronic warfare (EW) threats must be accounted for while planning the network deployment.
Closed-form analytical solutions commonly used for many optimization problems in the Operations Research community are not as useful because they cannot capture the network dynamics, which may frequently keep the system out of steady state operation. The dynamics may be caused, for instance, by the interplay of mobility and terrain effects (e.g., coverage in hilly regions), dynamic environmental factors (e.g., cloud coverage), bursty traffic (e.g. many users downloading video files), or potential loss of assets. Relying entirely on field measurements and live equipment to assess and mitigate performance gaps in a deployment quickly becomes cost-prohibitive.
Network Simulation-based optimizations, operating within the context of model-based systems engineering (MBSE), have been shown to be effective at bridging this gap. We believe that this is an area of significant growth for network planning tools and we are actively investing in this area in order to develop a set of automated tools to plan and optimize backhaul network laydowns for commercial and military customers.
NETWORK LAYDOWNS: Improving Network Performance Through Improved Laydowns
Given the expected mobility, traffic, and the technology to be used for the backhaul communications, the network planner should automatically generate initial and improved laydown to optimize coverage, throughput, or energy consumption. Heatmap based displays can be overlaid to visualize coverage (for instance, via relative signal strengths) together with detailed statistics.
With an established expectation around application traffic and type (text, video, voice), the laydown can be automatically refined to optimize bandwidth requirements and the heatmaps may also be used to show relative over-under utilization of the backhaul network assets. Subsequently, the laydown can be automatically improved, for instance, to reduce the number of towers needed to provide adequate coverage and bandwidth for a specified scenario. The optimization algorithms use both position and traffic information to refine the tower placements until specified objective(s) are satisfied by the laydown. The operator may also make manual updates to tower locations and other scenario attributes and view the impact of proposed changes on overall performance. The use of Machine Learning (ML) and AI-based technologies can be used to continuously improve coverage and respond to anticipated events, potentially by deploying temporary relays and similar assets.
THE SCALABLE SOLUTION
We at SCALABLE believe that network simulations embedded within, or operating in conjunction with, next generation network management and network operations suites, can play a central role in enhancing the ability of networks to respond effectively to dynamic or unanticipated changes in their operational conditions.
For more information on the SCALABLE Networks solution, visit us here.
About the Author
Dr. Rajive Bagrodia is the Founder and CEO of Scalable Network Technologies, Inc. and an Emeritus Professor of Computer Science at UCLA. Previously, Dr. Bagrodia served as a Professor of Computer Science at UCLA, where he led a research group in mobile computing and parallel and distributed programming that produced simulation systems such as Maisie, Parsec, and GloMoSim. His research was supported by large, multi-investigator grants from federal agencies including DARPA and NSF. Dr. Bagrodia founded SCALABLE Network Technologies in the wake of significant innovations his research group achieved in the theory and practice of performance prediction for complex, large-scale computer and communication systems. Today, SCALABLE is recognized as a global leader in the development of advanced simulation technology and in its application to enhance cyber resilience of commercial and military systems. Dr. Bagrodia received a Bachelor of Technology in Electrical Engineering from the Indian Institute of Technology, Bombay and a PhD degree in computer science from the University of Texas at Austin. He has published over 175 research papers in Computer Science journals and at international conferences on high performance computing, wireless networking, and parallel simulation.