Search for a command to run...
Blockchain technology offers solutions to numerous network problems by leveraging distributed record-keeping and collaborative decision-making features. However, deployment considerations such as blockchain infrastructure cost, performance requirements, and scalability are often overlooked. This paper provides an in-depth perspective on deploying blockchain-based solutions for telecommunications networks, estimating costs, comparing infrastructure options (on-premises, IaaS, BaaS), and choosing a suitable blockchain platform. We have analyzed prominent use cases and investigated deployment options, highlighting the pros and cons of each. Finally, we present two case studies, one proposing a distributed marketplace solution for 5G slice brokering and another one on the decentralization of federated learning (FL) through blockchain. Experiments are conducted to identify the performance limitations of the proposed solution under various deployment infrastructures. For the slice brokering use case, we studied the achievable transaction throughput and average latency under various systems under test with different resource specifications. Our experiments showed that while use cases that required maximum transaction throughputs in the range of 10 to 200 could be carried out with sub-second latency, use cases that require higher transaction throughputs (300 to 400) would need more computational resources to maintain such low latency. The federated learning use case provided insights into the achievable accuracy of distributed learning under various blockchain settings (public, consortium, and private). This led to the understanding that private and consortium blockchains can achieve acceptable accuracy in significantly lower training times compared to public blockchains.