Skip to main content Accessibility help
×
Home
  • Print publication year: 2013
  • Online publication date: July 2013

Preface

Summary

Why we wrote this book

Back in the late 1990s, when CDMA was widely considered the dominant technology for cellular 3G, two of the authors and a few colleagues in Bell Laboratories designed an alternative technology called Flash-OFDM with two basic yet fundamental ideas: OFDMA-based airlink and all IP-based network architecture. In early 2000, we founded a startup company, Flarion Technologies, to prove Flash-OFDM in the market by building terminals and base stations, and testing and deploying the networks in a wide variety of locations, configurations, and frequency bands. As arguably the first commercially deployed OFDMA/IP-based cellular system, Flash-OFDM helped make those two ideas the key enabling features in 4G mobile broadband LTE.

From the remarkable journey of designing, developing, and deploying Flash-OFDM, we have learned, and in some cases “unlearned,” a few important lessons:

• While early cellular wireless communications design focuses predominantly on the physical layer, mobile broadband requires more system-level thinking across different protocol layers than just the physical layer. For example, OFDMA, in comparison with CDMA, more readily facilitates a simplified IP-based network architecture design, where air interface specific technology functions and processing are collapsed into a base station and IP layer protocols are used for handoff.

Conventional wisdom developed in early cellular wireless communications needs to be reexamined from first principles. For example, the wireless channel is conventionally modeled with additive noise and multiplicative channel response; we found that selfnoise should also be included when multiplexing users with large signal dynamic range in OFDMA. As another example, universal frequency reuse is conventionally considered the most spectrally efficient; we found that for data, fractional frequency reuse improves both cell edge and cell average performance.