Receiver / PHY Technologies
The core foundation of TrellisWare’s receive processing is firmly grounded in the concept of using all the information available, all the time. This includes taking advantage of anything known about the environment and the likely operational usage for receiver optimization (e.g., for equalization, synchronization or interference mitigation) in addition to synergistically using the estimated demodulated and decoded data, sometimes referred to as “joint likelihood processing.” A high-level walkthrough of the basic ideas behind TrellisWare Likelihood Processing (TLP) is presented in the annotated briefing to the right.
Click to enlarge imageIf it’s known that a radio will be operating in a network of devices all using the same signal structure to communicate (as is common in commercial cellular or military networked radios), the structure of these possibly interfering signals should be used to optimize the receiver design. Similarly, if it’s known that the receiver will be moving quickly or the received environment will be changing abruptly (such as when the receiver rounds a corner or a new set of data is applied), this knowledge should be used to optimize the receiver’s design.
While TLP approaches can dramatically improve receive processing, they require deep expertise to implement effectively. If the correct trellis size and estimation techniques are not used, techniques like PSP can result in catastrophic performance failure, or equally as bad, extremely large silicon implementations. Each application must be considered uniquely, along with the processing partitioning and the “firing schedules” for iteration in AID; all play a critical role in how well the final solution performs and how difficult it is to implement.
TrellisWare engineers have more than a decade of experience building and testing these implementations, so they’re able to avoid often-subtle pitfalls and optimize each implementation for a given set of processing resources.
