Multi-h modulation indexes selection using the PBIL algorithm

Abstract

Multi-h CPFSK Modulation is a scheme that is simple to implement, spectrally efficient in terms of bandwidth and has inherent error correction properties [I, 21. It is the purpose of this research to find specific limits of the niche in which multi-h finds itself.The vehicle of comparisons between modulation and coding schemes is the Probability of Error tradeoff against signal to noise ratios. Multiti requires large parameter sets to obtain low Probabilities of Error. Huge numbers of combinations may need to be investigated. A fast efficient and thorough search needs to be performed. In this paper, an search algorithm based on the genetic algorithm and known as t,he 'Population Based Incremental Learning' (PBIL) Algorithm[S] is thought to be ideal. I INTRODUCTIONThere are an increasing number of applications requiring digital communications today, each demanding various resources for required levels of performance. These resources consist of the physical hardware and demands placed on the communications channel. The amount of hardware needed generally depends on the complexity of the modulation and encoding schemes chosen. The channel places spectral bandwidth and power restrictions so as to permit multiple users of the same channel.The channel user must be able to use an economical bandwidth which will allow for a required level of performance specified by the Probability of Error ai? a function of signal to noise ratio. This will lead to a choice of a particular modulation scheme that will ultimately be used.Constant-envelope continuous-phase signal modulation (CPM) has received much attention in applications of digital communications over satellite and terrestrial radio channels [4, 51. The reason is that CPM is a power and bandwidth efficient scheme and shows resistance to fading and nonlinear distortions [SI. Non-continuous phase shifts in the transmitted signal have numerous harmonics. CPM is bandwidth efficient due to it's continuous nature. The power spectrum of CPM has side bands which decrease in amplitude far more rapidly than non-continuous modulation schemes and therefore has a much smaller bandwidth requirement. CPM has a wider main spectral lobe than other competing schemes, and thus much more energy will be transmitted in the frequency range of this lobe in comparison to other schemes. As this is close to the desired transmission frequency, this lobe will pass narrower bandwidth restrictions unscathed. This is a more energy efficient way of transmitting signals.As CPM is phase modulation scheme, it will be fairly resiliant to noise effecting amplitude. Transmissions in channels arc subjcct to AWGN (Additive White Gaussian Noise) , Fading and Nonlinear distortions. Futhermore coding like convolutional encoding can provide further resistance to noise by incorporating an effective redundancy or memory which the channel does not have.In the family of CPM, there is digital scheme known as Continuous Phase Frequency Shift-Keying ( CPFSK). Amongst the group of CPFSK schemes, very strong codes allowing f...