Analysis of a Chirp-Based Waveform for Joint Communications and Radar Sensing (JC&S) using Non-Linear Components Andre N. Barreto, Thuy M. Pham, Sandra George, Padmanava Sen and Gerhard Fettweis Barkhausen Institut, Dresden, Germany Email: {andre.nollbarreto, minhthuy.pham, sandra.george, padmanava.sen, gerhard.fettweis}@barkhauseninstitut.org Abstract—Joint communications and radar sensing (JC&S) is expected to be one of the key features in beyond 5G (B5G) networks, allowing the provision of radar as a service (RaaS). In this paper, we are interested in a chirp-based waveform that can be effectively employed for both communication and radar applications. More specifically, we investigate the performance of such a waveform in the presence of realistic non-linear power amplifiers (PA) and low-noise amplifiers (LNA), operating at mmWave frequencies. I. I NTRODUCTION Sensing is widely seen as one of the key technologies in the sixth generation (6G) of wireless communication systems [1], allowing a plethora of new services and applications. It is envisaged that radar sensing, in particular, will be integrated with the communications network [2], such that it can be offered as an additional service upon demand, in what we call Radar as a Service (RaaS). Currently, radar and wireless communication systems are designed and deployed separately, using different hardware and waveforms, and distinct parts of the spectrum. Consumer radar applications, especially for the automotive industry, currently operate mostly in the 24 GHz and in the 76- 81 GHz bands, with the former being discontinued soon [3]. Anyway, there will still be 5 GHz of spectrum being used for automotive and consumer-market radar, not to mention the spectral allocation for aeronautical, meteorological and military radar. This is nearly twice as much as the whole 5G spectrum currently available, including the frequencies above 6 GHz. Radar services are already extensively used in vehicles, and their usage is likely to increase with new autonomous vehicles, but this usage is currently limited to highways and streets. Other usages of radar can be envisaged, like gesture recognition [4], but in most places, like homes, offices and parks, this sizeable portion of the spectrum remains largely unexploited. There have been some attempts in the literature to allow the coexistence of both systems [5] in the same spectrum band using cognitive-radio techniques. This approach can be effective when we have an incumbent primary service at fixed locations, but it is unlikely to make full usage of the available spectral resources in a more dynamic scenario. Also, in this approach, radar and communications are still considered as separate systems, using different waveforms and equipment. However, it is common knowledge that both applications rely on the same physical phenomenon, i.e., the propagation of electromagnetic waves, and, therefore, both radar sensing and wireless communications could be co-designed, such that they can share the same waveform, spectrum and hardware. This latter approach allows a flexible allocation of resources, depending on the temporal and spatial needs of each service, which results in a more efficient usage of the available spectrum. For instance, in a busy street crossing, most of the resources can be allocated to radar, whereas at home the spectrum can be fully allocated for communications, to allow communications at very high data rates. The service demands may also dynamically vary over time. For example, a radar ap- plication may consist of several distributed radars in different vehicles and in the infrastructure, which communicate with each other through an in-band broadband communications link, for interference coordination and/or sensor fusion. One of the research challenges towards this vision is how to design a physical layer (PHY) that is flexible, but still efficient for both radar and communications. Orthogonal frequency- division multiplexing (OFDM) has been proposed as a possible waveform [6], but its implementation can be rather complex for the large bandwidths under consideration, and would also require full duplexing terminals for monostatic radars. In fact, a chirp-based waveform can be effectively used to achieve both functionalities. Chirps are extensively used in radar systems, because of their good ambiguity properties, but mostly because of the possibility of a less-complex hardware implementation, using pulse compression and lower-rate sam- pling [7]. The performance of chirp signals for radar detection has been extensively studied in the literature and in text books [8], [9], and it will, therefore, not be addressed in this paper. Chirps can also be used for communications, and this is the focus of this paper. Frequency-shift keying (FSK) mod- ulated chirps have been proposed in [10], but, due to radar requirements, the data rate is limited by the chirp duration, which is usually much larger than the inverse of the bandwidth, resulting in a low spectral efficiency. The modulation with phase-shift keying (PSK) or quadrature-amplitude modulation