Time Division Multiple Access Methods for Wireless Personal Communications

David D. Falconer, Fumiyuki Adachi, and Björn Gudmundson

One-line summary: TDMA and various modulation and detection optimizations are presented.

Overview/Main Points

Basics: Bitstream broken into frames, frames broken into slots and slots are assigned to users. Upstream (user to base) and downstream (base to user) are duplexed with in time domain (TDD) or frequency domain (FDD). Slots contain data bits, CRC check bits, guard bits, synchronization training bits, and control bits.
Pros and Cons: Variable bit rate to users, less power control needed because slots are temporally separated within cell, and frequency separated amongst cells. Users can use idle slots to measure neighboring BS power levels for handoff (MAHO - mobile assisted handoff). Pulsating power envelope is bad - interference and complexity. High data rates needed as 1/N "duty cycle" for N users - equalization required to overcome delay spread.
ACA: adaptive channel allocation is an optimization in which frequencies are borrowed from less crowded cells to increase a cell's capacity. ACA also performed to dynamically avoid interference.
FH: frequency hopping is another optimization in which all transmitters in a cell cycle through a sequence of frequencies (offset from each other). Differet cells have random sequences, averaging out inter-cell interference. FH-TDMA is effectively spread spectrum CDMA with the sequence as code.
HCS: hierarchical cell structures are used to increase coverage - large macrocells fill in "holes" of microcells, and are used in MAHO. Macro and microcell frequencies are obviously orthogonally partitioned.
Modulation: QPSK (4-level modulation) best for wireless/cellular environments. GMSK and DQPSK also used, as is QAM. Coherent versus noncoherent modulation depends on time variability of the channel. Differential detection is the common form of noncoherent modulation, and can be improved with decision feedback of the past L symbols.
More optimizations: Convolutional coding and adaptive equalization can further combat multipath effects like delay spread.

Relevance

TDMA is a well-understood, commonly employed, and efficient media access technique. The enhancements described in this paper are worth understanding, but for the most part are "magic tricks" used to improve performance, from the perspective of the network layer and above.

Flaws

The main thrust of the paper is clear, but it tended to get sidetracked and bogged down in poorly explained technical details.

Back to index