This week we made some progress with the receiver implementation, adding a Picoblaze core to handle three main tasks: the serial protocol, for PC-side communication; the line protocol, for modem-to-modem communication; and the Automatic Gain Control (AGC) algorithm, for input gain control on the receiver board. The core can only handle 1024 instructions, but we think this will be more than enough to satisfy our requirements. The use of a soft core simplifies some of the implementation (since we can have more robust protocols with no extra area cost) as well as the potentiometer control.

For now, the serial protocol has been completely specified and implemented. The PAR timing analysis results show very little slack due to high load on the Picoblaze I/O address port. We will try to minimise this in order to have more freedom with the HDL description of the upstream processing blocks.

Due to the firmware bugs in the hydrophone we used to perform the underwater experiments, as described last week, we scheduled a second round of tests for Friday, this time around at sea or in the marina. We hope this represents a more realistic environment with less multipath, so we can better observe the receiver behaviour and measure the channel response.