Communication systems for the Internet of Things (IoT) are critical for the transfer of information between IoT devices and the Cloud. While cellular systems, such as 5G, can provide excellent connectivity for IoT there are many application areas where a ”one size fit all approach” may not be possible. For example unique channels present in underground and underwater environments cannot be properly handled by existing cellular systems. Providing energy sources for IoT devices to communicate is also another key challenge. The large number of IoT devices present in the environment will also require efficient and cost effective multiple access schemes. In this thesis physical layer communication techniques to overcome these communication challenges in IoT systems are proposed. The...[ Read more ]

Communication systems for the Internet of Things (IoT) are critical for the transfer of information between IoT devices and the Cloud. While cellular systems, such as 5G, can provide excellent connectivity for IoT there are many application areas where a ”one size fit all approach” may not be possible. For example unique channels present in underground and underwater environments cannot be properly handled by existing cellular systems. Providing energy sources for IoT devices to communicate is also another key challenge. The large number of IoT devices present in the environment will also require efficient and cost effective multiple access schemes. In this thesis physical layer communication techniques to overcome these communication challenges in IoT systems are proposed. These techniques include 1) adaptive delay spread OFDM for mobile waveguide channels, 2) non-coherent energy-detection MIMO Receivers and 3) analog precoding using highly reconfigurable antennas (HRA).

Adaptive delay spread OFDM for mobile waveguide channels has important communication applications in IoT when robots or vehicles with sensors are traveling inside waveguide channels such as tunnels or pipelines. The delay spread in these mobile wideband waveguide channels varies significantly across subcarriers because propagation occurs in modes whose characteristics change significantly with frequency. Therefore cyclic prefixes need to be unnecessarily long and this reduces throughput. My contribution to solve this problem is to use knowledge of the delay spread in each subcarrier to adaptively select and avoid subcarriers with delay spread exceeding a given cyclic prefix length. With this approach it is possible to create an OFDM system that is more efficient than systems satisfying the delay spread of all subcarriers. The results could be important in applications where vehicles operate in tunnels or where robots operate to detect defects in buried underground pipe distribution networks where water, oil or gas is flowing.

Energy efficiency and energy harvesting are also key issues in IoT applications. Non-coherent Energy-detection MIMO receivers use only the magnitudes or energy of the received signals and knowledge of the magnitudes of the channel fading gains. The receivers can be used in applications such as energy harvesting, distributed antenna systems (DAS) or massive MIMO where existing energy detectors can be leveraged for communication or where either coherent systems are impossible or require extensive fronthaul. My contribution extends previous non-coherent approaches to MIMO configurations. Proposed structures include, maximum likelihood, and extensions to Zero-forcing, MMSE and OSIC and their performance in flat Rayleigh fading channel with multiple amplitude shift keying (M-ASK) are investigated. The results could be important in developing hybrid RF energy harvester and communications systems that exploit the rate-energy tradeoff and help make long life IoT devices possible.

Analog precoding using highly reconfigurable antennas (HRA) makes use of new developments in antenna design to arrive at implementable techniques for analog precoding for massive MIMO multiuser applications. HRA can generate various radiation patterns without any phase shifter and has been proven to be implementable. My contribution is to develop methods for using HRA in analog precoding. The feasibility and performance of HRA to replace antenna arrays to conduct analog precoding is investigated. The techniques could have important applications in the development of efficient and cost effective massive MIMO access points in serving IoT devices.

In all the investigations presented, performance evaluation and comparison with existing techniques is performed using either analytical or numerical techniques. Furthermore, where possible, experimental results are also provided. In addition conclusions and suggestions for further research are also included in the thesis.