Ho and Valaee extend the notion of mobile hotspots to mass transportation systems. They first explain why the conventional approach of applying the notion of access points of wireless local area networks (WLANs) is not feasible, and explore the challenges faced when employing cellular systems in truly mobile environments. The paper also considers the use of microcells along the transportation route, and explains why these are not feasible because of the high mobility of the transportation systems. Low earth orbit (LEO) satellite services are also ruled out due to service cost and portability. The paper thus makes the case for the authors’ system, which is applicable to both WLAN and cellular systems.

The proposal employs an open architecture that consists of repeaters located along the trackside and multiple antennas mounted on the exterior of the vehicle. This is described in section 2. Two methods of transmission between the repeater-antenna pairs are suggested and explained. The blind information raining method is introduced and analyzed in section 3. Improvement on system reliability with erasure coding on packets is also explained. The interference model used between the set of active repeaters and antennas is presented in section 4. In section 5, the authors propose and explain a throughput-optimized method based on resource allocation. Power, rate, and signal-to-noise ratios, together with the freedom of matchings between repeaters and antennas, are considered for resource allocation problems. The mathematical complexity of the problem is shown to be nondeterministic polynomial time (NP) complete. Section 6 presents several matching algorithms to show that a high normalized aggregate link rate from multiple links may be reached when power allocation algorithms are processed with the matchings. Based on the matching results, several power allocation algorithms are developed and explained in section 7. Section 8 presents a simulation of a typical scenario that reveals approximate values for some of the variables of the model. The simulation shows how some of the parameters of the modeled railway environment affect performance in terms of system throughput. In section 9, the authors present their conclusions.

This paper is strongly recommended for anyone involved in networking, wireless communications, and mobile communication systems. It is very well written, and is a welcome contribution to the discipline.