In contrast to most wireless communication systems which employ digital signal processing(DSP) only at baseband, systems with software radio usually start DSP at an intermediate frequency(IF)band. By using programmable DSP chips at IF as well as at baseband, software radio systems are very flexible and can efficiently support multi-band and multi-standard communications. The input to an IF stge of a software radio receiver is in general a very wideband signal, which is converted into a digital by bandpass sampling. The purpose of DSP at this stage is to isolate the signal of interest, which is usually a very narrowband signal, from a wideband input and to translate the signal down to baseband. Because the symbol rate of each standard is differnet, timing recovery problem is very important to support multi-standard communication. This paper propose new interpolation schemes for timing recovery for TDMA systems with software radio. These schemes reduce MSE of interpolated symbol by cascading noise suppression filter and conventional interpolation filter or cascading conventional interpolation filter and RLS filter. Using simulations for BPSK and QPSK systems in AWGN and Rayleigh fading channels, the performance of the proposed algorithm is verified. Also the computational complexity for hardware implementation is compared.