LLMs as Reasoning Engines for System Design
Definition: LLMs as Reasoning Engines
LLMs as Reasoning Engines
LLMs can assist with system design decisions by reasoning about trade-offs. Example: "Given constraints X, Y, Z, which modulation scheme minimizes BER while meeting throughput requirements?"
Example: System Design with LLM Reasoning
Use chain-of-thought prompting to design a MIMO-OFDM system.
Solution
Structured Prompt
prompt = '''Design a MIMO-OFDM system for the following scenario:
- Indoor office, 100 users, 1 Gbps aggregate throughput
- Available bandwidth: 100 MHz at 28 GHz
Think step by step about: antenna config, subcarrier spacing,
channel model, modulation, coding.'''