E-PoW: Haɗa Koyo na AI da Hakar Blockchain a Tsarin 6G

Teburin Abubuwan Ciki

80%

Ƙarfin Lissafi da aka Ceto

90%

Aikin AI daga MMC

6G

Tsarin da aka Yi Nufi

1. Gabatarwa

Haɗa fasahar Hankali na Wucin Gadi (AI) da blockchain a cikin tsarin zamani na shida (6G) yana gabatar da damammaki da ƙalubale. Yayin da AI ke ba da damar hanyar sadarwa mai hankali da bincike na bayanai, blockchain yana tabbatar da tsaro da bayyana gaskiya. Duk da haka, horar da AI yana buƙatar albarkatun lissafi masu yawa, waɗanda suka iyakance a cikin na'urorin 6G, kuma tsarin Shaidar Aiki (PoW) na al'ada na blockchain yana cinye ƙarfin lissafi mai yawa don ayyukan hakar ma'adinai, wanda galibi ake zarginsa da ɓarna.

2. Bayanan Baya da Ayyukan Da suka Danganci

2.1 Tsarin 6G da Bukatun AI

Ana hasashen tsarin 6G don tallafawa aikace-aikacen AI na ko'ina da ke buƙatar ayyukan lissafin matrix mai yawa. Bisa ga binciken Google akan Na'urorin Sarrafa Tensor, kusan kashi 90% na aikin AI yana fitowa daga fahimi mai yawa da hanyoyin sadarwa masu maimaitu, dukansu sun dogara sosai akan lissafin ninka matrix (MMC).

2.2 Hanyoyin Yarjejeniya na Blockchain

Yarjejeniyar PoW ta al'ada ta ƙunshi masu hakar ma'adinai suna yin bincike mai ƙarfi don neman ƙimar hash da aka yi niyya, suna cinye makamashi mai yawa. Madadin yarjejeniyoyi kamar Shaidar Raba (PoS) da Shaidar Aiki (PoA) suna rage yawan amfani da makamashi amma suna iya lalata rarrabuwa da tsaro.

3. E-PoW: Ingantaccen Shaidar Aiki

3.1 Tsarin Fasaha

E-PoW yana haɗa lissafin matrix daga horon AI cikin tsarin hakar blockchain. Hanyar yarjejeniya tana ba masu hakar ma'adinai damar yin lissafin AI mai mahimmanci yayin da suke neman tubalan da suka dace a lokaci guda, yana haɗa koyo na AI da hakar blockchain ta hanyar raba albarkatun lissafi yadda ya kamata.

3.2 Tushen Lissafi

Babban ƙirƙira yana cikin haɗa ayyukan matrix cikin tsarin hakar ma'adinai. An sake tsara matsalar hakar ma'adinai don haɗa da tabbacin ninka matrix:

$H(block\_header || nonce || MMC\_result) < target$

Inda $MMC\_result = A \times B$ yake wakiltar lissafin ninka matrix daga ayyukan horon AI.

Algorithm na Hakar E-PoW

function ePowMine(block_header, AI_tasks):
    while True:
        nonce = generate_random_nonce()
        
        # Yi lissafin matrix na AI
        matrix_result = compute_MMC(AI_tasks)
        
        # Lissafin haɗe hash
        hash_input = block_header + nonce + matrix_result
        hash_value = sha256(hash_input)
        
        if hash_value < target_difficulty:
            return (nonce, matrix_result, hash_value)
        
        update_AI_tasks()

4. Aiwatarwa da Sakamako

4.1 Saitin Gwaji

An gwada yarjejeniyar E-PoW a cikin yanayin kwaikwayon 6G tare da nodes masu hakar ma'adinai da yawa waɗanda ke yin ayyukan horon AI a layi daya ciki har da rarraba hotuna da samfuran sarrafa harshe na halitta.

4.2 Bincike na Aiki

Sakamakon gwaji ya nuna cewa E-PoW zai iya ceto har zuwa kashi 80% na ƙarfin lissafi daga hakar tubalan tsantsa don horon AI a layi daya. Tsarin ya kiyaye tsaron blockchain yayin da yake ƙara saurin haɗuwar samfurin AI.

Kwatanta Aiki: E-PoW da PoW na Al'ada

Bayanin Ginshiƙi: Taswira ginshiƙi da ke nuna kwatanta raba albarkatun lissafi tsakanin E-PoW da PoW na al'ada. E-PoW yana nuna kashi 80% na albarkatun an keɓe su don horon AI da kashi 20% don hakar ma'adinai, yayin da PoW na al'ada yana nuna kashi 100% an keɓe su don hakar ma'adinai ba tare da amfani da AI ba.

5. Aikace-aikacen Gaba

E-PoW yana da babbar yuwuwar a cikin yanayin lissafi na gefe, tsarin koyo na tarayya, da hanyoyin sadarwa na IoT inda ingantaccen lissafi ke da mahimmanci. Ci gaban gaba zai iya haɗawa da fasahohi masu tasowa kamar lissafin neuromorphic da tsarin blockchain masu jurewa ƙididdigewa.

Bincike na Asali

Yarjejeniyar E-PoW tana wakiltar sauyin yanayi a yadda muke fuskantar raba albarkatun lissafi a cikin tsarin rarrabuwa. Ta hanyar gane tushen lissafi na gama gari tsakanin horon AI da hakar blockchain, marubutan sun ƙirƙiri dangantakar haɗin gwiwa tsakanin fasahohi biyu da ake ganin sun banbanta. Wannan hanya ta yi daidai da ƙa'idodin da aka gani a cikin wasu tsare-tsaren lissafi na ƙirƙira, kamar tsarin CycleGAN (Zhu et al., 2017) wanda ya samo haɗin kai da ba a zata ba tsakanin yankuna daban-daban ta hanyar tsarin lissafi da aka raba.

Abin da ya sa E-PoW ya fi jan hankali musamman shi ne hanyarsa ta aiki ga sanannen matsala. Ba kamar yawancin shawarwarin ka'idoji da ke sadaukar da tsaro don inganci ba, E-PoW yana kiyaye kaddarorin tsaro da aka tabbatar na PoW na al'ada yayin da yake inganta ingancin lissafi sosai. Wannan ya yi daidai da binciken daga Shirin 6G na IEEE, wanda ke jaddada buƙatar hanyoyin yarjejeniya masu ingancin makamashi a cikin hanyoyin sadarwa na gaba.

Adadin ƙarfin lissafi kashi 80% da aka ceto da aka nuna a cikin gwaje-gwajen yana da ban mamaki, musamman idan aka yi la'akari da cewa wannan bai lalata ainihin kaddarorin blockchain ba. Wannan ribar inganci na iya yin tasiri mai zurfi ga ayyukan blockchain mai dorewa, yana magance ɗaya daga cikin manyan sukar hakar ma'adinan cryptocurrency. Hanyar tana kama da yadda tsarin TPU na Google ya inganta don ayyukan matrix waɗanda suka mamaye duka AI da wasu nau'ikan lissafin sirri.

Idan aka duba gaba, E-PoW zai iya ba da damar sabbin nau'ikan aikace-aikace a cikin hanyoyin sadarwa na 6G inda AI da blockchain dole ne su kasance tare yadda ya kamata. Kamar yadda aka lura a cikin ƙayyadaddun 3GPP don hanyoyin sadarwa na gaba, haɗa AI da fasahar rajistar rarrabuwa zai zama mahimmanci don ayyukan hanyar sadarwa mai cin gashin kanta. E-PoW yana ba da takamaiman hanyar aiwatarwa zuwa ga wannan hangen nesa.

Duk da haka, ƙalubale sun rage a cikin daidaita ayyukan lissafin matrix da tabbatar da gasa mai adalci tsakanin masu hakar ma'adinai tare da ƙarfin lissafi daban-daban. Aikin gaba yakamata ya bincika hanyoyin daidaita wahala masu dacewa waɗanda ke la'akari da duka wahalar hakar ma'adinai da lissafin AI, kama da yadda algorithms na binciken tsarin jijiya na zamani ke daidaita manufofi da yawa.

6. Bayanan da aka ambata

Zhu, J. Y., Park, T., Isola, P., & Efros, A. A. (2017). Unpaired Image-to-Image Translation using Cycle-Consistent Adversarial Networks. IEEE International Conference on Computer Vision.
Nakamoto, S. (2008). Bitcoin: A Peer-to-Peer Electronic Cash System.
King, S., & Nadal, S. (2012). PPCoin: Peer-to-Peer Crypto-Currency with Proof-of-Stake.
IEEE 6G Initiative. (2023). Roadmap to 6G: Connecting Everything by 2030.
3GPP Technical Specification Group. (2024). Study on Scenarios and Requirements for Next Generation Access Technologies.
Jouppi, N. P., et al. (2017). In-Datacenter Performance Analysis of a Tensor Processing Unit. ACM/IEEE 44th Annual International Symposium on Computer Architecture.