Abstract : Orthogonal Frequency Division Multiplexing (OFDM) has been employed as an attractive solution for wireless communication, because of the digital implementation of OFDM in which subcarrier generation and data modulation are accomplished digitally using orthogonal transformation such as Fast Fourier Transform (FFT) of data block. This technique is considered as a building block of the suggested spread spectrum scheme. A suggested scheme is defined as Direct Sequence Orthogonal Frequency Division Multiplexing (DS-OFDM) which combines both Direct Sequence (DS) and Frequency Hopping (FH). In this scheme, data are multiplied by Pseudo Noise (PN) sequence at rate much higher than the original bandwidth. This is done in frequency domain rather in time domain. The system operates at rate of 9600 bits/second using QPSK modulation technique, and with different framing length that contain two different data lengths over time varying HF channel. A comparison was established with simulated conventional DS system. The simulation results show that the conventional DS can not operate successfully in any one of the assumed channel conditions while the suggested scheme results in significant improvement especially in (poor, flutter) channel conditions.For example, gain of about 4 dB is recorded at BER of 10-4 in poor channel condition at the presence interference, further 2dB is resulted from the use of more frame length at the same BER. In contrast, Noticeable improvement can be obtained in flutter channel condition. The second part of this thesis is concerned with suggested error detection and correction algorithm which has the a ability to detectcorrect then synchronize the spreading sequence without the need to generate the same sequence at the receiver. This algorithm can be applied to the conventional DS systems. The implementation of the suggested algorithm shows that it can work over several channel conditions (Interference, Phase error) even it cost more iterations to reach the steady state (15 iterations at maximum in the applied channel conditions). From the above results and assuming the ability of detecting and correcting the received data using the suggested error detection and correction algorithm associated with advantage of OFDM properties, a suggested algorithm for channel impulse response estimation is implemented using maximum of 15 iterations from the suggested algorithm. For example, A mean square error of a bout -20 is obtained when the framing length is increased.