Joint equalization and decoding for two-dimensional intersymbol interference channels

Joint equalization and decoding schemes for channels with two-dimensional intersymbol interference are presented. Advanced storage technologies, such as patterned magnetic media, two-dimensional optical storage, and holographic storage, and multiuser communications scenarios, such as cellular communications or optical communication systems employing wavelength-division multiplexing, are examples of systems that have two-dimensional intersymbol interference. Two types of schemes are proposed: schemes that use conventional equalization techniques, like minimum mean-square-error equalization, in conjunction with low-density parity-check codes; and schemes that perform message-passing on a joint factor graph representing the channel and the low-density parity-check code. The schemes are iterative in nature; the equalizer and decoder exchange information and use it for subsequent iterations. The schemes are designed to incorporate the physics of the underlying problem. For the minimum mean-square-error based schemes, various low-complexity equalizers are proposed to study the trade-off between complexity and performance. The message-passing scheme is used for decoding on channels with nonlinear two-dimensional intersymbol interference; in particular, it is applied to the channel model of the two-dimensional optical storage system being developed by Philips, Netherlands. An anomalous behavior of the sum-product message-passing decoder is presented; the decoder performs better when fed wrong information about the channel, i.e ., introduction of channel mismatch is beneficial for the sum-product algorithm based decoder. The joint equalization and decoding schemes are applied to magnetic recording channels with inter-track interference. In this case the recording is simulated using a micromagnetic model for longitudinal magnetic recording. The use of array codes as error-correcting codes for the joint equalization and decoding schemes is also studied. A high-density data storage system employing an array of two-dimensional storage devices is proposed. For the same system; a concatenated coding scheme that uses the message-passing decoder as an inner code and array codes as outer codes is also proposed.