In this talk, we describe how convolutional codes can be used in a randomized coding set-up for physical layer security. We then utilize these codes to obtain randomized concatenated coding solutions - namely, randomized parallel (and serial) concatenated convolutional coding schemes. We develop the corresponding iterative low-complexity decoders, and evaluate the randomized codes' performance in terms of the resulting security gap. We demonstrate via several examples that the newly developed solutions can outperform other alternatives in the literature. For instance, when the proposed code concatenation is combined with scrambling, we can obtain a security gap of less than 1 dB for an eavesdropper error rate ~0.5 and main channel error rate 10^-5.