Shulou(Shulou.com)11/24 Report--

CTOnews.com, February 26 (Xinhua) A team of scientists led by the Massachusetts Institute of Technology has developed a decoder chip. The relevant research results were presented at the ongoing International Conference on solid State Circuits.

The chip uses a general decoding algorithm developed before MIT, and can decode any encoded signal. Decoding data is simpler and faster than traditional technology, and its energy consumption is only 1% to 10% of that of other similar hardware. in other words, the performance is 10 to 100 times higher than other hardware, and can be widely used in areas that are demanding on power consumption, such as virtual reality and 5G networks.

▲ picture: Muriel M é dard, a professor of Christine Daniloff science and NEC software science and engineering and professor of electrical engineering and computer science at the Massachusetts Institute of Technology, said, "this is the first time that someone has broken the decoding threshold of 1 picojoule per bit. This is roughly the same amount of energy you need to transfer 1bit within the system. This is a significant symbolic threshold, but it also changes the balance of the receiver, which is probably the most urgent part from an energy point of view-we can move it from the decoder to other parts.

In addition to M é dard, the co-authors of this paper include Arslan Riaz, a graduate student at Boston University; Rabia Tugce Yazicigil, an assistant professor of electrical and computer engineering at Boston University; Ken R. Duffy, then director of the Hamilton Institute at Menus University and now a professor at Northeastern University; and other scholars from the Massachusetts Institute of Technology, Boston University and Menus University.

CTOnews.com here is a simple science popularization, the common data is transmitted in the form of bits (0 or 1), and the sender needs to encode the data and add error correction codes at the end of the data. The error correction code mentioned here is also a redundant string composed of 0 and 1, which can be regarded as the information needed for hash verification. This string of information is usually stored in a specific codebook, and the decoding algorithm at the receiver is an authentication scheme designed for this special password, which will retrieve the original information that may have been confused through a specific codebook and hash structure. In general, each algorithm is based on specific code, and most algorithms require dedicated hardware, so a device needs many chips to decode different data.

The researchers previously demonstrated a general decoding algorithm that can crack any code, GRAND (guessing Random Additive noise Decoding), which works by guessing the noise that affects the transmission of information, then removing the noise directly from the received data, and then checking the rest of the codebook. It guesses a series of noise patterns in the order in which they may appear.

In fact, the data we receive usually contains reliability information, also known as soft information, which can help the decoder find out which parts are wrong.

It is reported that this new decoding chip, called ORBGRAND (Ordered Reliability Bits GRAND), can use this soft information to sort data according to the possibility of errors in each bit. But in fact, it's not as simple as sorting individual bits. While the most unreliable parts may be the most obvious mistakes, the third and fourth unreliable parts may be as wrong as part 7. Here, ORBGRAND uses a new statistical model to sort bits in this way, because multiple bits put together can cause the same error as a single bit.

"if your car breaks down halfway, the soft message may tell you that it is the battery. But it's troublesome if it's not just the battery, such as the battery and the engine at the same time, "says M é dard, and this is how a rational person does it-although it may be a failure of the two parts together, but you will start with the place where it is most likely to go wrong, and then look down to find something that is unlikely to happen.

They believe that this is a more effective method than traditional decoders. It is reported that traditional decoders only focus on code structure, and it is usually designed for the worst-case scenario. "with a traditional decoder, you have to take out the car's blueprints and then check every part, although you will eventually find the problem, but it will take you a long time," M é dard explained.

It is reported that as soon as a key code is found, ORBGRAND will stop sorting, which is usually very fast. In addition, the chip also uses parallel logic, which can generate and test a variety of noise patterns at the same time, so as to find this key code more quickly.

When they compared this method with other chips, the ORBGRAND decoded with the highest accuracy consuming only 0.76 picojoules per bit, breaking the previous performance / energy consumption record. This also means that ORBGRAND consumes only 1/10 or even 1% of the energy consumption of other devices.

M é dard said that one of the biggest challenges in developing chips is to reduce energy consumption. ORBGRAND is now so energy-efficient that other processes that researchers didn't pay attention to before, such as checking the codewords in the codebook, consumed most of the effort.

"now, the inspection process is like turning on the car to see if it works, which is the hardest part. So we need to find more effective ways to do this." the team is also exploring ways to change the transmission modulation so that they can use ORBGRAND chips to improve efficiency.

Welcome to subscribe "Shulou Technology Information " to get latest news, interesting things and hot topics in the IT industry, and controls the hottest and latest Internet news, technology news and IT industry trends.