The parameters of physical layer radio frame for 5th generation (5G) mobile cellular systems are expected to be flexibly configured to cope with diverse requirements of different scenarios and services. This paper presents a frame structure and design, which is specifically targeting Internet of Things (IoT) provision in 5G wireless communication systems. We design a suitable radio numerology to support the typical characteristics, that is, massive connection density and small and bursty packet transmissions with the constraint of low-cost and low complexity operation of IoT devices. We also elaborate on the design of parameters for random access channel enabling massive connection requests by IoT devices to support the required connection density. The proposed design is validated by link level simulation results to show that the proposed numerology can cope with transceiver imperfections and channel impairments. Furthermore, the results are also presented to show the impact of different values of guard band on system performance using different subcarrier spacing sizes for data and random access channels, which show the effectiveness of the selected waveform and guard bandwidth. Finally, we present system-level simulation results that validate the proposed design under realistic cell deployments and inter-cell interference conditions.INDEX TERMS 5G, frame structure, Internet of Things, random access channel.

In this paper, we present a novel random access method for future mobile cellular networks that support machine type communications. Traditionally, such networks establish connections with the devices using a random access procedure, however massive machine type communication poses several challenges to the design of random access for current systems. State-of-the-art random access techniques rely on predicting the traffic load to adjust the number of users allowed to attempt the random access preamble phase, however this delays network access and is highly dependent on the accuracy of traffic prediction and fast signalling. We change this paradigm by using the preamble phase to estimate traffic and then adapt the network resources to the estimated load. We introduce Preamble Barring that uses a probabilistic resource separation to allow load estimation in a wide range of load conditions and combine it with multiple random access responses. This results in a load adaptive method that can deliver near-optimal performance under any load condition without the need for traffic prediction or signalling, making it a promising solution to avoid network congestion and achieve fast uplink access for massive MTC.

No abstract

No abstract