Line Coding Technique for Channel Equalization: Integrated Pulse-Width Modulation and Consecutive Digit Chopping
Abstract : This paper presents two new line-coding schemes, integrated pulse width modulation (iPWM) and consecutive digit chopping (CDC) for equalizing lossy wireline channels with the aim of achieving energy efficient wireline communication. The proposed technology friendly encoding schemes are able to overcome the fundamental limitations imposed by Manchester or pulse-width modulation encoding on high-speed wireline transceivers. A highly digital encoder architecture is leveraged to implement the proposed iPWM and CDC encoding. Energy-efficient operation of the proposed encoding is demonstrated on a high-speed wireline transceiver that can operate from 10 to 18 Gb/s. Fabricated in a 65-nm CMOS process, the transceiver operates with supply voltages of 0.9 V, 1 V, and 1.1 V. With the help of the proposed iPWM encoding, the transceiver can equalize over 27-dB of channel loss while operating at 16 Gb/s with an efficiency of 4.37 pJ/bit. The design occupies an active die area of 0.21 mm2.
? Conventional feat techniques on the receiver finish admire call feedback equalizers have tight feedback temporal arrangement constraints, that end in higher power consumption because the rate increases.
? Feed forward equalization (FFE) on the transmitter with voltage mode driver avoids the feedback path and ends up in economical equalization.
? However, PWM encoding needs the insertion of a particular narrow pulse in each knowledge bit.
? These slender pulses should be accurately reproduced at the transmitter output, that necessitates very wide information measure within the high-speed data path, leading to poor energy potency and problem in scaling PWM coding to higher data rates.
? Although such a divided FFE implementation helps to keep up aconstant output termination resistance (50) across all faucet settings, it comes at the price of
(a) increased signal power,
(b) increased change power since multiple segments are needed to attain desired one-dimensionality, and
(c) tight coupling between fifty termination standardization and FFE tap coefficients tuning.
? These 3 constraints cut back the FFE potency because the range of FFE faucets are increased to equalize serious channel loss.
• The proposed CDC encoding is impressed by the magnetic recording systems, that introduced management transitions within the knowledge stream to cut back the impact of pulse state of affairs.
• A pulse of opposite polarity in the middle of CIDs introduces high-frequency element in the data stream, which helps to reduce the post-cursor ISI.
• Energy-efficient operation of the proposed encoding is incontestable on a high-speed wireline transceiver which will operate from ten to eighteen Gb/s.
• A highly-digital encoder design is proposed to implement the iPWM and government agency scheme, which is ascendible in nature and achieves widerange high-resolution secret writing constant tuning.
? The performance of the transceiver was improved further by using additional power in the receiver, to compensate a loss of 27dB.
? The performance summary of the proposed transceiver and compares it with state-of-the-art transceiver designs.
? Equalization of the channel loss consumes significant power and degrades the energy efficiency of the wireline communication link.
? Conventional line-coding techniques such as Manchester encoding (also known as pulse width modulation or PWM), can equalize the wireline channel without increasing signaling power, without segmenting the output driver, and without coupling the 50 termination tuning with the coefficient tuning.
|