OFDM Waveform Optimisation for Joint Communications and Sensing

Radar systems are radios to sense objects in their surrounding environment. These operate at a defined set of frequency ranges. Communication systems are used to transfer information between two points. In the present day, proliferation of mobile devices and the advancement of technology have led to communication systems being ubiquitous. This has made these systems to operate at the frequency bands already used by the radar systems. Thus, the communication signal interferes a radar receiver and vice versa, degrading performance of both systems. Different methods have been proposed to combat this phenomenon. One of the novel topics in this is the RF convergence, where a given bandwidth is used jointly by both systems. A differentiation criterion must be adopted between the two systems so that a receiver is able to separately extract radar and communication signals. The hardware convergence due to the emergence of software-defined radios also motivated a single system be used for both radar and communication. A joint waveform is adopted for both radar and communication systems, as the transmit signal. As orthogonal frequency-division multiplexing (OFDM) waveform is the most prominent in mobile communications, it is selected as the joint waveform. Considering practical cellular communication systems adopting OFDM, there often exist unused subcarriers within OFDM symbols. These can be filled up with arbitrary data to improve the performance of the radar system. This is the approach used, where the filling up is performed through an optimisation algorithm. The filled subcarriers are termed as radar subcarriers while the rest as communication subcarriers, throughout the thesis. The optimisation problem minimises the Cramer--Rao lower bounds of the delay and Doppler estimates made by the radar system subject to a set of constraints. It also outputs the indices of the radar and communication subcarriers within an OFDM symbol, which minimise the lower bounds. The first constraint allocates power between radar and communication subcarriers depending on their subcarrier ratio in an OFDM symbol. The second constraint ensures the peak-to-average power ratio (PAPR) of the joint waveform has an acceptable level of PAPR. The results show that the optimised waveform provides significant improvement in the Cramer--Rao lower bounds compared with the unoptimised waveform. In compensation for this, the power allocated to the communication subcarriers needs to be reduced. Thus, improving the performances of the radar and communication systems are a trade-off. It is also observed that for the minimum lower bounds, radar subcarriers need to be placed at the two edges of an OFDM symbol. Optimisation is also seen to improve the estimation performance of a maximum likelihood estimator, concluding that optimising the subcarriers to minimise a theoretical bound enables to achieve improvement for practical systems