Spectrally-temporally adapted Spectrally Modulated Spectrally Encoded (SMSE) waveform design for coexistent CR-based SDR applications

Spectrally Modulated, Spectrally Encoded (SMSE) waveforms have demonstrated considerable practical utility and remain viable alternatives for implementing Cognitive Radio (CR) techniques in Software Defined Radio (SDR) applications. A key benefit of CR-based SDR platforms is their potential for alleviating spectrum scarcity by efficiently exploiting temporal-spectral regions that are under-utilized. When operating under limited bandwidth constraints and amid dissimilarly structured coexisting signals, CR-based SDR signals can be designed such that they "peacefully" coexist while maintaining "manageable" levels of mutual interference in other systems. In this research, the goal is to expand applicability of the SMSE framework by developing a waveform optimization process that enables intelligent waveform design. The resultant waveforms are capable of adapting to a spectrally diverse transmission channel while meeting coexistent constraints.;SMSE waveform design is investigated with respect to two different forms of coexisting signal constraints, including those based on resultant interference levels and those based on resultant power spectrum shape. As is demonstrated, the SMSE framework is well-suited for waveform optimization given its ability to allow independent design of spectral parameters. This utility is greatly enhanced when soft decision selection and dynamic assignment of SMSE design parameters are incorporated. Results show that by exploiting statistical knowledge of primary user (PU) spectral and temporal behavior, the inherent flexibility of the SMSE framework is effectively leveraged such that SMSE throughput (Bits/Sec) is maximized while limiting mutual coexistent interference to manageable levels. This process is accomplished using independent selection of subcarrier modulation order and power allocation. Additional gains are achieved by accounting for the temporal behavior of coexistent signals, thereby allowing the SMSE system to statistically predict optimal waveform designs. Results demonstrate an approximate 20% increase in throughput is achieved by employing a Reactive Spectrally-Temporally adapted waveform design relative to a Spectrally-Only adapted design, with an additional 10% increase in throughput realized using a Predictive Spectrally-Temporally adapted design.;SMSE system capability is extended further using uniform spectral partitioning with carrier-interferometry (CI) coding to increase SMSE waveform agility. By adaptively varying the modulation order and optimally allocating power within each spectral partition, inherent SMSE flexibility is more fully exploited and SMSE throughput substantially increases in the presence of spectral mask constraints. Results demonstrate up to a 36% increase in throughput is achieved by employing spectral partitioning, with up to 110% improvement achieved by employing spectral partitioning in conjunction with a Predictive Spectrally-Temporally adapted waveform design.