Benefits and Challenges of Deterministic Reference Channel Models

Abstract

The paper introduces a new paradigm for reference channel models. Current reference channel models are designed as platforms that generate radio channels for testing using random values for their parameters. These parameters follow some pre-established distribution based on process called parameterization, i.e. statistical processing of previous real measurements or accurate ray tracing simulations. The paper argues that random generated channels give either no new insight or even delusive information and should be replaced with the initial set of radio channels that was used for parameterization. Therefore a deterministic reference channel model, as an emulator of previously recorded real radio channels, is proposed and its potential elaborated.