University of California

Fred Tzeng

Payam Heydari

Amin Jahanian

A code-division multiplexing (CDM) system utilized in multi-channel (MC) front-end integrated circuits to significantly reduce the power consumption of such systems. The CDM system extends data compression advantages to uncorrelated and weakly correlated MC signals through the introduction of a new Multi-Channel Signal Binning and Multiplexing (MCSBM) method and architecture. The method achieves significant reductions in power consumption in comparison to a conventional time-division multiplexing quantizer, while adding only a modest amount of overhead and complexity. Systems and methods permit architects to fabricate MC integrated circuits with ultra low power consumption and small chip area. Another embodiment relates to the system's compressor organizing samples of the input signal in such a way that the downstream analog-to-digital converter quantizes the higher variance samples with a higher resolution compared to the resolution it uses to quantize other samples with lower variance.

Multi-channel code-division multiplexing in front-end integrated circuits

A method of adaptively and losslessly quantizing an analog signal to a digital signal in an analog-to-digital converter (ADC), is disclosed. According to one embodiment, the quantizing is based on one or more of an instantaneous amplitude of the analog signal, frequencies of the analog signal, and patterned contents of the analog signal, and the method comprises sampling the analog signal; quanitizing the analog signal by a quantizer core, wherein the quantizer core comprises a digital-to-analog converter (DAC), a comparator, and a voltage reference, wherein the quantization is one of efficient lossless and adaptive compression quantization, or a traditional quantization method; determining by an analog compression engine (ACE) whether the sampled analog signal is above or below a defined threshold, wherein the defined threshold includes one or more of an amplitude threshold, a frequency threshold, and a patterned adaptive threshold; adaptively manipulating the sampled signal mathematically through at least one of the DAC or voltage reference of the quantizer core; and outputting the digital signal.

Signal statistics and compression-based analog-to-digital converters

A challenging problem is maintaining low power consumption in multi-channel (MC) systems, where multiple input signals demand several front-end analog signal-processing blocks to be replicated. A code-division multiplexing (CDM) system can be generalized as a signal compression-decompression involving an analog-to-digital converter (ADC). This invention utilizes CDM in MC front-end integrated circuits to significantly reduce the power consumption of such systems. The invention also extends CDM's data compression advantages to uncorrelated and weakly correlated MC signals through the introduction of a new Multi-Channel Signal Binning and Multiplexing (MCSBM) method and architecture. The proposed method achieves significant reductions in power consumption in comparison to a conventional time-division multiplexing quantizer, while adding only a modest amount of overhead and complexity. Among other advantages, the invention permits architects to fabricate MC integrated circuits with ultra low power consumption and small chip area, where conventional architectures could not. Another embodiment of the invention relates to the system's compressor organizing samples of the input signal in such a way that the downstream ADC quantizes the higher variance samples with a higher resolution compared to the resolution it uses to quantize other samples with lower variance.

Described herein are code-modulated multi-signal systems. In one embodiment, a multi-signal system receives multiple input signals and code-modulates each input signal with a unique code to distinguish the input signal from the other input signals. The input signals may come from multiple antennas, multiple sensors, multiple channels, etc. The code-modulated signals are then combined into a combined signal that is sent through shared blocks and/or transmitted across a shared medium in a shared path. After shared processing and/or shared transmission, the individual signals are recovered using matched filters. Each matched filter contains a code corresponding to one of the unique codes for recovering the corresponding signal from the combined signal. The recovered signals may then be inputted to additional processors for further processing.

Code-modulated path-sharing multi-signal systems

Described herein are multi-band LNAs that reuse inductors for different frequency bands to minimize chip area. In an embodiment, a multi-band LNA is capable of operating in a narrowband (NB) and a wideband (WB) while reusing at least one input impedance matching inductor and at least one load inductor for both bands. The reuse of inductors results in a more efficient use of chip area. In an exemplary embodiment, the LNA comprises a common source transistor and a common gate transistor. In this embodiment, the LNA operates in a common source configuration using the common source transistor to amplify input signals in the NB, and operates in a common gate configuration using the common gate transistor to amplify input signals in the WB. The LNA reuses an input impedance matching inductor and a load inductor in both configurations, and thus both bands.

Growing community of inventors

Cerritos, CA, United States of America

Fred Tzeng