Comprehensive Energy Model of Wireless Transceiver Front Ends
As CMOS technology scales down, digital supply voltage and digital power consumption goes down. However due to dynamic range limitations, power supply and power consumption of the RF front-ends and analog sections do not scale in the same fashion. In fact, in scaled systems, the RF section of a wireless transceiver consumes more energy than the digital part. For better understanding of the design trade offs, we first develop an accurate and comprehensive energy model for the analog front-end of wireless transceivers. Next, we evaluate a single user point-to-point wireless data communication system and a multi-user CDMA based system with respect to RF front end energy consumption and communication quality. We demonstrate the effect of occupied signal bandwidth, peak-to-average ratio (PAR), symbol rate, constellation size, and pulseshaping roll-off factor on single user system, and the effect of number of users and multiple access interference (MAI) on CDMA based multi-user system. For a given quality specification, we show how the energy consumption can be reduced by adjusting one or more of these parameters. Wireless communication and mobile computing devices are widely used in everyday life. All of these devices are powered by a limited lifetime battery source. Since the advances in battery technology have failed to keep up with the battery capacity demands in mobile communications, aggressive techniques to reduce the power consumption of wireless communication devices have to be developed. Different aspects of low power wireless communication have been addressed in recent years. These include the effect of modulation scaling [1], lazy packet transmission [2], energy efficient routing [3] etc. In these power analyses, the RF circuit energy is often ignored or over simplified. However, the RF section processes analog signal with higher frequency and consumes more energy than the digital part. To analyze RF circuit energy consumption, an accurate and comprehensive RF energy model is necessary. Several transceiver energy models have been proposed in recent years. For microsensor systems, the transceiver energy model considers the circuit start-up energy in addition to steady state dissipated energy in [4]. The energy consumption of every component is assumed to be constant. Another high level model was proposed in [1], which divides the transceiver circuitry power into two parts, one related to the instantaneous symbol rate and the other to the highest symbol rate. This model is not comprehensive since the dissipated circuit power is related not only to the symbol rate, but also to parameters, such as peak-to-average ratio (PAR), required resolution in data converters, signal bandwidth, sampling frequency, etc. The energy model in [5] is more comprehensive but it also assumes a constant power value for most components and the power model for power amplifier is only gain dependent. In this paper, we first present an accurate and comprehensive energy model for the RF front-end of a wireless transceiver. The components include ADC, DAC, reconstruction filter, mixer, frequency synthesizer, power amplifier, low noise amplifier (LNA) and baseband amplifier. We consider the effects of signal bandwidth, PAR, symbol rate, sampling frequency, constellation size, etc. on the power consumption of each of these components. We next study the role of PAR on the RF front-end energy consumption and symbol error rate (SER) for both single user and multi-user wireless communication systems. We show the effect of symbol rate, thermal noise, roll-off factor and the constellation sizes on the performance of a single user system. We also demonstrate the effect of MAI, roll-off factor and number of users on a CDMA based multi-user system. The remainder of this paper is organized as follows. Section 2 describes the transceiver system model and defines the terminology used in this paper. Section 3 describes the power model for every component in the wireless transceiver. Section 4 describes the effect of design parameters such as PAR and signal bandwidth on the RF front-end energy consumption. Section 5 analyzes the communication quality of both single user wireless data transmission and multi-user CDMA system. Section 6 summarizes the paper.