Expressions for the cutoff rate of memoryless channels and certain channels with memory are derived assuming decoding with approximate channel statistics. For channels with memory, two different decoding techniques are examined: conventional decoders in conjunction with ideal interleaving/deinterleaving, and maximum likelihood decoders that take advantage of the channel memory. As a practical case of interest, the cutoff rate for the band-limited nonlinear satellite channel is evaluated where the modulation is assumed to be M-ary phase shift keying (MPSK). The channel nonlinearity is introduced by a limiter in cascade with a traveling wave tube amplifier (TWTA) at the satellite repeater while the channel memory is created by channel filters in the transmission path.

We evaluate and compare bit error rates for conventional (linear and nonlinear) repeater satellites and regenerative satellites. This work is based on a new formulation for error rates using the generalized moment technique for numerical evaluation of error rates. Both coherent MPSK and noncoherent MFSK modulations are considered with regenerative satellites. The results show that regenerative satellites are especially effective against uplink CW interference when on-board despreading of spread spectrum signals is included as part of the satellite processing.

A solution to the mean time to lose lock for a first order, pseudo-noise, code tracking loop is derived for both the fulltime delay-lock loop and the time-shared delay lock loop, with both utilizing an early-late gate size of 1/2 chip, and assuming ideal bandpass filters. The result allows a spread spectrum system designer to assess the slip time performance of his code tracking loop by a simple calculation and then a reading of a curve presented in this paper. The results presented here are actually very general and apply to other code loop spacings and types. It also applies to carrier tracking loops and agrees with the carrier loop result of Viterbi [3].