An efficient congestion control scheme for a lightweight constrained application protocol in the Internet of Things

The Internet of Things is a promising technology which tends to revolutionize and connect the global world via heterogeneous smart devices through seamless connectivity. The current demand for machine-type communications has resulted in a variety of communication technologies with diverse service requirements to achieve the modern Internet of Things vision. More recent cellular standards like long-term evolution have been introduced for mobile devices but are not well suited for low-power and low data rate devices such as the Internet of Things devices. To address this, there is a number of emerging Internet of Things standards. Fifth-generation mobile networks, in particular, aim to address the limitations of previous cellular standards and be a potential key enabler for the future Internet of Things. Additionally, the third-generation partnership project has introduced low-power wide-area cellular-based networks such as extended coverage global systems for mobile communications for the Internet of Things, enhanced machine-type communications and narrowband-Internet of Things as enabling solutions to support the new service requirements for massive to critical Internet of Things use cases. Therefore, in a comprehensive literature review, this study highlights the state-of-the-art application requirements of the Internet of Things, along with their associated emerging and enabling communication technologies, with the main focus on fifth-generation mobile networks that are envisaged to support the exponential traffic growth for enabling the Internet of Things. The study further investigates the challenges and open research directions pertinent to the deployment of a massive-critical Internet of Things in coming up with a context-aware congestion control in a Constrained Application Protocol for resource-constrained devices. A profound open research challenge from the literature review is the need for a context-aware congestion control (CACC) approach for a lightweight CoAP/UDP-based Internet of Things traffic. The CACC proposes mechanisms that include retransmission timeout (RTO) estimator, retransmission count based smoothed round-trip-time (RTT) observation, lower-bound RTT restriction approach, and aging concept. The proposed RTO estimators utilise the strong, weak, and failed RTT to identify the exact network status and provide adaptive congestion control. The CACC incorporates a variable of the retransmission count in the request-response interaction model to mitigate the negative variation in RTT due to the fluctuations in the Internet of Things environment. Moreover, with a lower-bound RTO restriction approach, the unnecessary spurious retransmissions are avoided, and the aging mechanism limits the validity of the RTO value to improve the efficiency of the proposed scheme. The proposed CACC model is validated against baseline CoAP and CoCoA+ using Contiki OS and the Cooja simulator. The results obtained are impressive under different network scenarios. Managing congestion control in a resource-constrained lossy network with a high bit error rate is a challenging task that needs to be given due consideration if the ever-growing promises of the Internet of Things are to be actualised. The primary congestion control mechanism defined by the core CoAP specification is not capable of adapting to the bursty traffic conditions. This calls for and motivates the need for further research in congestion control mechanisms. The study proposes a congestion control scheme that utilises a Particle Swarm Optimisation (PSO)-based Adaptive Congestion control Technique (PACT). PACT applies random and optimal parameter-driven simulations to optimise the default CoAP parameters in order to adapt to the traffic conditions. The PSO-based algorithm varies the retransmission and max-age values for different traffic scenarios. The proposed PACT exhibits significant performance in terms of packet loss, delay, and normalised overhead in comparison with baseline CoAP with Observe under different network and traffic scenarios.Thesis (PhD (Electronic Engineering))--University of Pretoria, 2020.Electrical, Electronic and Computer EngineeringPhD (Electronic Engineering)Unrestricte