The analytical model evaluates the performance of grant-free communications in a dense IoT environment

Imagine a world where all smart devices can communicate seamlessly, from traffic sensors to wearable health monitors. This vision is at the heart of large-scale machine type communications (MMTC), the basis for 5G and future 6G mobile networks.

Simply put, MMTC aims to connect an unprecedented number of Internet of Things (IoT) devices to one million people per square kilometre. This ability is important for advances in smart cities, self-driving cars, and remote healthcare, among many other applications.

An important approach to enabling this large-scale connectivity is the “grant-free” communications scheme. Unlike traditional cellular communications, where a device must first request permission to send from the base station, grant-free schemes allow devices to send data without prior permission.

This simplifies the communication process, significantly reduces end-device processing and power consumption, and reduces base station scheduling operations. However, there are significant drawbacks to the Grant-Free scheme. When many devices are transmitted simultaneously, there is a higher risk of data collisions, leading to network congestion and communication failures.

To tackle these key challenges, a research team led by Professor Shino of the Graduate School of Informatics at the University of Chiba, Japan has developed a comprehensive analytical model to assess the performance of grant-free communication schemes. Their paper, published in Computer Communications, explores how the widely known, grant-free method known as “slotted aloha” works in a densely populated IoT environment.

Other members of the team include Yuki, a graduate of Chiba University, and Professor Takeshima, a professor at the Graduate School of Information Science and Technology at Osaka University.

This paper is an expanded version of the research that won the Best Paper Award at ACM MSWIM 2023 (presented at the 26th International Conference on Modeling, Analysis and Simulation of Wireless and Mobile Systems), an international conference in the field of performance modeling and evaluation of wireless communication systems ranked in the Core Conference Rank A.

The team’s approach involves creating sophisticated analytical models based on probabilistic geometry, a mathematical tool for analyzing systems using randomly distributed elements. They assumed that both base stations and IoT devices were scattered throughout the region in a statistically random yet predictable way.

We then analyzed three different scenarios of slotted aloha. A basic version with no special enhancements, a version that incorporates an interference cancellation technique called “NOMA,” and a version that uses power controls that allow the device to adjust its own signal strength. The team focused on two key performance indicators: transmission success probability and base station throughput (the amount of data that a base station can successfully receive in a given time frame).

Their findings revealed many of the complex dynamics of different aloha versions. Interference cancellation improved the base station throughput by up to 20%, but in some cases it improved by up to 20%, but did not resolve what was called a “near-night problem.” This is a phenomenon that devices close to the base station are far more likely to have successful transmissions, and more than that is fighting.

Surprisingly, this study found that interference cancellation is most effective for devices with intermediate distances, not very close or very far from the base station. On the other hand, applying power controls successfully addressed almost surrounding issues and ensured a more fair transmission opportunity for all devices, but led to a significant reduction in overall network performance.

“Our research reveals that aloha-based communication faces an inherent trade-off between two conflicting objectives. Equity is the goal of allowing a single base station to receive data from as many devices as possible, in the sense that devices can communicate regardless of their distance from the base station, and in the sense that they are throughput.

“In other words, achieving both fairness and maximum throughput simultaneously is fundamentally difficult.” This highlights the key design challenges for future IoT networks. This suggests that relying solely on grant-free schemes may be infeasible to achieve both optimal performance and fair access.

Overall, the results of this study will help guide the development of IoT. Understanding the fundamental trade-offs of communication schemes is important for designing efficient and fair next-generation networks.

“We have shed light on the inherent limitations of IoT networks that form the backbone of future IoT-driven societies. These limitations can be overcome by adopting future work, due to the use of grant-free communications schemes. We intend to further explore this possibility.

Such exciting applications in society include vehicle and road infrastructure exchange data, and vehicle-to-vehicle communication using remote medical care using wearable devices.