This document discusses techniques for improving the reliability of Network-on-Chip (NoC) designs. It begins by explaining the importance of fault tolerance in NoCs due to increasing technology scales. It then describes different types of faults and provides an overview of current reliability techniques including error correction codes, retransmission mechanisms, reliable task mapping, and fault-tolerant routing. Specific schemes for self-healing routers, error detection, power analysis, and resilience against negative bias temperature instability are also summarized. The document concludes by stating that while these techniques improve reliability, most increase power consumption, and future work should focus on reducing this overhead through thermal-aware designs and methods to selectively wear out cores.