Cooperative and non-cooperative Sense-and-Avoid (SAA) systems are key enablers for Unmanned Aircraft (UA) to routinely access non-segregated airspace. In this paper the stateof-the-art cooperative and non-cooperative SAA technologies for small size UA are presented and the associated multisensor data fusion techniques are discussed. The non-cooperative sensors including both passive and active Forward Looking Sensors (FLS) and the cooperative systems including Traffic Collision Avoidance System (TCAS), Automatic Dependent Surveillance - Broadcast (ADS-B) system and/or Mode C transponders form part of the SAA holistic architecture. After introducing the SAA system processes, the key mathematical models are presented. The Interacting Multiple Model (IMM) algorithm is used to estimate the state vector of the intruders and this is propagated to predict the future trajectories using a probabilistic model. Adopting these mathematical models, conflict detection and resolution strategies for both cooperative and un-cooperative intruders are identified. Additionally, a detailed error analysis is performed to determine the overall uncertainty volume in the airspace surrounding the intruder tracks. This is accomplished by considering both the navigation and the tracking errors affecting the measurements and translating them to unified range and bearing uncertainty descriptors, which apply both to cooperative and non-cooperative scenarios. Detailed simulation case studies are carried out to evaluate the performance of the proposed SAA approach on a representative host platform (AEROSONDE UA) and various intruder platforms, including large transport aircraft and other UA. Results show that the required safe separation distance is always maintained when the SAA process is performed from ranges in excess of 500 metres.