A Large Set of Secure Signatures for DS-SS Long Range Radars in Autonomous Cars

The current trend in the use of ICT to boost the development of the automotive sector is mainly ad- dressed towards so called autonomous vehicles, equipped with appropriate sensors, actuators and processors, that make them able to move safely, without the intervention of a human driver. Actually, since several years, ICT permitted a widespread implementation of systems to help the driver maintaining the control of the vehicle, even when exceeding its normal limits of handling. Among them, vehicle radars may work under all weather conditions, provide a satisfactory coverage, and see several meters ahead. This paper investigates the performance, in terms of detection capability and false detection rate, of Direct Sequence Spread Spectrum Long Range Radars adopt- ing binary De Bruijn sequences as radar signatures, compared to more classical solutions adopting Gold codes. Following the analysis of the correlation and security properties of De Bruijn sequences, the radar detection per- formance are discussed in two typical scenarios, and by means of specific tools, such as the ambiguity function to account for Doppler effects. Even if De Bruijn sequences do not provide disruptive improvements with respect to Gold codes, they exhibit better behavior in the presence of Doppler, and a much greater cardinality of their set, at a parity of the length, which may be useful to accommodate many potential users in a vehicular scenario.