Search

The primary requirements of a secure Wireless Sensor Network architecture are confidentiality, integrity and authentication of users and other participating entities. User Authentication for wireless sensor networks is a fundamental and important issue in designing dependable and secure systems. In this thesis, we have outlined the security model, functional requirements, assumptions and network setup for an authentication scheme in the first phase. Keeping in mind the security requirements as well as the flaws of past authentication schemes, we propose a robust user authentication method that inherits user anonymity, mutual authentication and password changing functionality of previous password-based schemes and improves security by resisting gateway bypass and replay attack, and many logged in user with the same ID threat. Our scheme is a variant of strong password based schemes that does not require strict network synchronization. In the second phase of the thesis, we have analysed our authentication scheme from the perspective of security issues and functional requirements. The proposed scheme is modelled in SystemC. It is evaluated in different attack scenarios. The authentication latency, memory and functional requirements, and computational overhead are the metrics used to evaluate the scheme. The effect of multiple users on authentication latency in our scheme is also studied. Some of the past representative schemes have also been modelled and evaluated in the same environment. A detailed comparison of over-head cost, authentication latency and security features are provided in this thesis. It is verified and confirmed by modeling that our scheme provides enhanced security without adding extra computation at the sensor node.

The 802.11ac version of the popular IEEE 802.11 protocol aims to boost performance by increasing the channel bandwidth and allowing simultaneous transmission to multiple clients in the downlink direction. This dissertation presents an innovative approach of performance evaluation of downlink multi-user multiple input multiple output (DL-MUMIMO) technique for differentiated quality of service (QoS) based traffic categories in IEEE 802.11ac protocol. We propose a novel analytical model based on discrete Markov chain (DMC) and E-limited M/G/1 queuing model to evaluate the performance improvements of multi-user transmission opportunity (MU-TXOP) sharing in DL-MU-MIMO under non-saturated load and non-ideal channel condition. We also evaluate MU-TXOP sharing probabilities among different QoS nodes to assess the performance improvement of different traffic categories. In this dissertation, we propose an Access Point controlled MAC protocol (A-MAC) that enables simultaneous transmissions from multiple stations (STA) in uplink to eliminate under utilization of network resources in uplink transmission due to single user communication. The protocol uses enhanced distributed channel access (EDCA) technique to initiate multi-user transmission and orthogonal frequency division multiple access (OFDMA) method to transmit multiple Ready-To-Send (RTS) messages simultaneously. The proposed protocol also introduces explicit channel sounding technique by using dedicated OFDM subcarrier blocks for each user. We evaluate the performance of the proposed protocol using a noble analytical model and validate the performance metrics by extensive simulation in different traffic conditions. In this dissertation we also propose the increase of carrier sensing threshold (CSTH) of nodes during association with access point (AP) to mitigate collision probability due to hidden nodes during uplink transmission and validate our proposal through extensive simulation. Finally, we propose a noble analytical model to evaluate the performance of restricted access window (RAW) mechanism of IEEE 802.11ah as the MAC layer protocol for internet of things (IoT) network. We evaluate detailed performance metrics of non-QoS IoT network and investigate the feasibility of using RAW mechanism to support differentiated QoS based heterogeneous IoT network