CN-122029546-A - Fixed quantization alignment in machine learning based channel state information compression

Abstract

Systems, methods, apparatuses, and computer program products for fixed quantization alignment in machine-learning based Channel State Information (CSI) compression. The method may include training, at a network device, at least one automatic encoder with a quantized perceptual or quantized non-perceptual scheme, the at least one automatic encoder having at least one fixed quantization setting, and determining quantized block information, the quantized block information being determined based on the at least one fixed quantization setting. The method may further include transmitting encoder model information to the user equipment and obtaining channel state information by dequantizing the received compressed channel state information using at least one fixed quantization setting.

Inventors

• S. Rezai
• XING YUNCHOU
• F. Tosato
• CHEN JIE

Assignees

• 诺基亚技术有限公司

Dates

Publication Date

20260512

Application Date

20240809

Priority Date

20230811

Claims (20)

1. 1.A method, comprising: Training, at a network device, at least one automatic encoder with a quantized perceptual or quantized non-perceptual scheme, the at least one automatic encoder having at least one fixed quantization setting; determining quantization block information, the quantization block information being set based on the at least one fixed quantization; Transmitting encoder model information to a user equipment, and Channel state information is obtained by dequantizing the received compressed channel state information using the at least one fixed quantization setting.
2. 2. The method of claim 1, wherein the obtaining channel state information further comprises decoding the channel state information based on the received compressed channel state information that is dequantized.
3. 3. The method of claim 1 or 2, further comprising: And sending the quantized block information to the user equipment.
4. 4. The method of claim 1 or 2, further comprising: the quantized block information is received from the user equipment.
5. 5. The method of any one of claims 1 to 4, further comprising: transmitting an indication to the user equipment to indicate enablement of fixed quantization, and An acknowledgement of the capability to support the fixed quantization is received from the user equipment.
6. 6. The method of any of claims 1 to 5, wherein the encoder model information comprises at least one of: Encoder model structure or encoder model weights.
7. 7. The method of any of claims 1 to 5, wherein the encoder model information includes training channel state information and corresponding potential vectors based on the at least one fixed quantization setting.
8. 8. The method of any of claims 1 to 7, wherein the quantization block information comprises the fixed quantization settings.
9. 9. The method of any of claims 1-7, wherein the quantization block information includes a selected level and codebook for the at least one fixed quantization setting.
10. 10. A method, comprising: At the user device, establishing a quantization block based at least in part on the quantization block information; at the user device, establishing an encoder based at least in part on encoder model information received from a network device; estimating channel state information of a channel; compressing the channel state information of the channel using the quantization block and the encoder, and The compressed channel state information is sent to the network device.
11. 11. The method of claim 10, further comprising: Receiving an indication from the network device to enable fixed quantization, and An acknowledgement of the capability to support the fixed quantization is sent to the network device.
12. 12. The method of claim 10 or 11, wherein the encoder model information comprises at least one of: Encoder model structure or encoder model weights.
13. 13. The method of claim 10 or 11, wherein the encoder model information includes training channel state information and corresponding potential vectors based on the at least one fixed quantization setting.
14. 14. The method of any of claims 10 to 13, wherein the quantized block information comprises at least one fixed quantization setting.
15. 15. The method of any of claims 10 to 13, wherein the quantization block information comprises a selected level and codebook for the at least one fixed quantization setting.
16. 16. An apparatus, comprising: at least one processor, and At least one memory storing instructions that, when executed by the at least one processor, cause the apparatus to at least: Training at least one automatic encoder with a quantized perceptual or quantized non-perceptual scheme, the at least one automatic encoder having at least one fixed quantization setting; determining quantization block information, the quantization block information being set based on the at least one fixed quantization; Transmitting encoder model information to a user equipment, and Channel state information is obtained by dequantizing the received compressed channel state information using the at least one fixed quantization setting.
17. 17. The apparatus of claim 16, wherein the obtaining channel state information comprises the at least one memory storing instructions that, when executed by the at least one processor, cause the apparatus to decode the channel state information based on the received compressed channel state information that is dequantized.
18. 18. The apparatus of claim 16 or 17, wherein the at least one memory stores instructions that, when executed by the at least one processor, cause the apparatus to: And sending the quantized block information to the user equipment.
19. 19. The apparatus of claim 16 or 17, wherein the at least one memory stores instructions that, when executed by the at least one processor, cause the apparatus to: the quantized block information is received from the user equipment.
20. 20. The apparatus according to any of claims 16 to 19, wherein the at least one memory stores instructions that, when executed by the at least one processor, cause the apparatus to: transmitting an indication to the user equipment to indicate enablement of fixed quantization, and An acknowledgement of the capability to support the fixed quantization is received from the user equipment.

Description

Fixed quantization alignment in machine learning based channel state information compression Technical Field Some example embodiments may generally relate to mobile or wireless telecommunication systems, such as Long Term Evolution (LTE) or fifth generation (5G) radio access technologies or New Radio (NR) access technologies, or other communication systems. For example, certain example embodiments may relate to apparatus, systems, and/or methods for fixed quantization alignment in machine-learning based Channel State Information (CSI) compression. Background Examples of mobile or wireless telecommunications systems may include Universal Mobile Telecommunications System (UMTS) terrestrial radio access network (UTRAN), LTE evolved UTRAN (E-UTRAN), LTE-advanced (LTE-a), multeFire, LTE-a Pro, fifth generation (5G) radio access technology or NR access technology, and/or 5G-advanced. The 5G wireless system refers to the Next Generation (NG) wireless system and network architecture. The 5G network technology is mainly based on NR technology, but a 5G (or NG) network may also be built on top of the E-UTRAN radio. It is estimated that NR can provide bit rates of about 10-20 Gbit/s or higher and can support at least enhanced mobile broadband (eMBB) and ultra-reliable low-delay communications (URLLC) as well as large-scale machine type communications (mMTC). NR is expected to provide extremely broadband and ultra-robust, low latency connections, and large scale networks to support the internet of things. Mobile and wireless telecommunication systems often communicate channel state information. Channel state information is sometimes compressed prior to delivery over a mobile or wireless network to reduce overhead of communications. Disclosure of Invention One embodiment may be directed to an apparatus. The apparatus may include at least one processor and at least one memory storing instructions. The instructions stored in the at least one memory may, when executed by the at least one processor, cause the apparatus at least to perform training at least on an automatic encoder with a quantized perceptual or quantized non-perceptual scheme, the at least one automatic encoder having at least one fixed quantization setting, and determining quantized block information. The quantization block information may be based on at least one fixed quantization setting. The instructions may also cause the apparatus to transmit encoder model information to the user equipment and obtain channel state information by dequantizing the received compressed channel state information using at least one fixed quantization setting. Another embodiment may be directed to an apparatus. The apparatus may include at least one processor and at least one memory storing instructions. The instructions stored in the at least one memory may, when executed by the at least one processor, cause the apparatus to at least perform establishing a quantization block based at least in part on the quantization block information. The instructions may also cause the apparatus to establish an encoder based at least in part on encoder model information received from the network device, and estimate channel state information for the channel. The instructions may also cause the apparatus to compress channel state information for the channel using the quantization block and the encoder, and include transmitting the compressed channel state information to the network device. One embodiment may be directed to a method. The method may include training, at a network device, at least one automatic encoder with a quantized perceptual or quantized non-perceptual scheme, the at least one automatic encoder having at least one fixed quantization setting, and determining quantized block information. The quantization block information may be based on at least one fixed quantization setting. The method may further include transmitting encoder model information to the user equipment and obtaining channel state information by dequantizing the received compressed channel state information using at least one fixed quantization setting. Another embodiment may be directed to a method. The method may include establishing, at a user device, a quantization block based at least in part on quantization block information. The method may include establishing, at a user device, an encoder based at least in part on encoder model information received from a network device, and estimating channel state information for a channel. The method may also include compressing channel state information for the channel using the quantization block and the encoder, and transmitting the compressed channel state information to the network device. Another embodiment may be directed to an apparatus. The apparatus may include means for training at least one automatic encoder with a quantized perceptual or quantized non-perceptual scheme, the at least one automatic encoder having at least one fixed quantization setting,