Performance of military communication systems is nowadays improving slowly compared to commercial systems which puts interests in evolving mature commercial systems for military usage. In parallel, Public Safety (PS) communication systems are changing due to emergence of Long Term Evolution (LTE) as a mature solution to replace the legacy ones while providing new services. However, LTE is initially designed for commercial cellular network and need to be further evolved to tackle the substantial requirements of PS use cases. For instance, opportunistic deployments require modifications to enable the autonomous operation and meshing of moving base stations while satisfying heterogeneous frequency band availability. In this thesis, we develop a complete solution to answer constrained PS and military use-cases allowing to wirelessly mesh mobile network nodes and to provide access to standard user equipment while only requiring a single radio LTE band.
Starting from PS and military communication systems landscape and use-cases, we present the potential scenarios and derive functional requirements for future wireless communication systems to allow new services and coverage of these use-cases. We further underline the specific constraints applying to these communication systems due their specific environment of use, especially limited availability of frequency resources. This leads to the selection of LTE as the RAT for both backhaul and access of the wireless system. We then review current LTE state of the art and LTE systems and compare them against these requirements. Underlining the limitations of such systems, we detail the challenges faced by a LTE solution that would answer these requirements. Then, we present a novel network infrastructure architecture that enables multi-hop LTE mesh networking for nomadic and autonomous base stations via in-band self-backhauling relying on a new base station: the e2NB. We detail the building blocks of the architecture to answer the multiple challenges. Specifically, we investigate the coordination and orchestration functionality within the proposed architecture and propose a cross layer hierarchical resource scheduling algorithm in order to efficiently meet Quality of Service (QoS) requirements for real-time traffic while maximizing the throughput for elastic flows. To demonstrate the feasibility and reliability of our proposed architecture, we implement the corresponding self-backhauling air interface based on Open Air Interface platform and compare with the legacy LTE air-interface. We then evaluate the efficiency and adaptability of our proposed resource scheduling algorithm in various network topologies and heterogeneous traffic flows with QoS requirements. Finally, we summarize the remaining uncertainties concerning real-field deployments and we conclude this study.