Digital Log-to-Linear Gain Conversion for Low-Bandwidth RFIC Links
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The communication channel between a digital baseband controller and a radio frequency integrated circuit (RFIC) in wireless communication devices has low bandwidth, making it inefficient to convert logarithmic gain to linear gain due to complex inverse-log computations, leading to degraded signal fidelity and higher error rates due to prolonged gain change times.
Innovation Solution
A method and system that convert logarithmic values into linear values using a coarse gain block to shift and a fine gain block to multiply digital input signals, generating a digital output signal efficiently, thereby reducing bandwidth requirements and minimizing silicon area and clock rate consumption.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If the digital baseband controller converts logarithmic gain to linear gain before transmission, then the RFIC can apply the gain correctly, but the conversion requires complex inverse-log computation and many bits to be transmitted over multiple transactions
Solution Approach 1:
The patent segments the gain value into two parts: a coarse gain portion and a fine gain portion. The coarse gain is transmitted using fewer bits, and the fine gain is transmitted using additional bits. This segmentation allows the gain to be updated more quickly by first applying the coarse gain and then refining it with the fine gain, thereby reducing the overall gain transition time while maintaining accuracy.
Solution Approach 2:
The patent applies preliminary action by first transmitting and applying the coarse gain portion before transmitting the fine gain portion. This preliminary application of gain reduces the time required for the overall gain transition, as the majority of the gain change is achieved quickly with the coarse portion, and the fine portion serves as a refinement.
2Measurement precision
If the digital baseband controller transmits linear gain values with a range of approximately 100 dB, then the RFIC can apply the correct gain, but many transactions are needed to complete the transfer due to low bandwidth
Solution Approach 1:
The patent segments the gain value into a coarse gain portion and a fine gain portion, allowing the majority of the gain information to be transmitted in fewer bits initially. This segmentation reduces the number of transactions needed to transmit the complete gain value, thereby improving productivity while maintaining the full precision of the gain control through the addition of the fine gain portion.
3Measurement precision
If the gain changes over a greater period of time due to multiple transactions, then all gain bits are transmitted accurately, but the receive signal fidelity is significantly degraded resulting in higher error rates
Solution Approach 1:
The patent segments the gain update into a coarse gain portion and a fine gain portion, allowing the signal to be adjusted more quickly with the coarse gain. This reduces the time during which the signal is in an intermediate state, thereby improving reliability and reducing error rates. The fine gain portion is then applied to achieve the final precise gain value, maintaining both speed and accuracy.
4Measurement precision
If conventional methods are used to convert logarithmic gain to linear gain, then the correct gain can be applied, but complex inverse-log computation is required
Solution Approach 1:
The patent segments the gain conversion process into two simpler operations: converting the coarse gain portion and converting the fine gain portion. This segmentation avoids the need for a single complex inverse-log computation over the entire gain range, thereby reducing device complexity while maintaining gain accuracy through the combination of both portions.
Data Source
AI summary
A method for converting logarithmic values into linear values in digital logic is provided. The method includes receiving a logarithmic gain signal. A shift signal is generated based on the logarithmic gain signal. A remainder signal is generated based on the logarithmic gain signal. A linear gain is applied to a digital input signal based on the shift signal and the remainder signal to generate a digital output signal.