The global telecommunications landscape is standing on the precipice of a fundamental architectural shift. As the industry moves beyond the initial deployment phases of 5G and begins to define the parameters of 6G, the integration of Artificial Intelligence (AI) into the very fabric of the Radio Access Network (RAN) has transitioned from a theoretical concept to a strategic imperative. In a landmark announcement that signals the next phase of this evolution, Nokia and SoftBank Corp. have formalized an agreement to conduct a large-scale commercial AI-RAN field trial in Japan, scheduled for 2027.<\/p>\n
This collaboration aims to validate the performance, scalability, and economic viability of AI-native networks under real-world commercial traffic conditions. By leveraging NVIDIA\u2019s accelerated computing prowess alongside Nokia\u2019s software-defined RAN expertise and SoftBank\u2019s proprietary orchestration platforms, the trial represents one of the most ambitious attempts to date to merge the worlds of high-performance computing (HPC) and mobile telecommunications.<\/p>\n
The 2027 field trial is not merely a laboratory demonstration but a rigorous assessment of how AI-RAN can function within a live, high-density commercial network. The project is built upon a sophisticated technology stack involving three primary stakeholders: Nokia, SoftBank, and NVIDIA.<\/p>\n
The trial will utilize Nokia\u2019s anyRAN<\/strong> software, a cloud-native solution designed to offer deployment flexibility across various hardware environments. This software will be integrated into SoftBank\u2019s AITRAS<\/strong> (AI-RAN Architecture) platform. AITRAS serves as the core infrastructure for SoftBank\u2019s vision of a network that can simultaneously process radio signals and AI workloads.<\/p>\n At the heart of the hardware layer lies the NVIDIA Aerial RAN Computer (ARC)<\/strong>. This platform provides the necessary computational "muscle" to handle the intensive mathematical requirements of both 5G\/6G signal processing and generative AI applications. To manage these disparate workloads, the companies will employ the AITRAS Orchestrator<\/strong>, a specialized management tool designed to balance resource allocation between telecommunications functions and AI processing tasks.<\/p>\n The primary objective of the 2027 trial is to move AI-RAN from "proof-of-concept" to "commercial readiness." Key focus areas include:<\/p>\n The journey toward the 2027 field trial is the result of years of incremental innovation and strategic alignment between the involved parties.<\/p>\n During this period, SoftBank and Nokia began exploring the limitations of traditional, hardware-centric RAN. As 5G adoption scaled, it became clear that the power consumption and static nature of traditional base stations were ill-suited for the dynamic needs of the burgeoning AI economy. Nokia launched its "anyRAN" approach to decouple RAN software from specialized hardware, while SoftBank began internal R&D on the AITRAS platform.<\/p>\n Early in 2024, the industry saw the formation of the AI-RAN Alliance<\/strong> at Mobile World Congress (MWC) Barcelona. Both SoftBank and Nokia (along with NVIDIA) were founding members. This alliance was created to foster an ecosystem where AI and wireless communications could co-evolve. Throughout 2024, Nokia and SoftBank conducted preliminary lab tests in Japan and Finland, proving that NVIDIA GPUs could successfully accelerate RAN workloads.<\/p>\n In the years leading up to the 2027 trial, the companies are focused on deep software integration. This involves "tuning" Nokia\u2019s anyRAN to work seamlessly with NVIDIA\u2019s CUDA ecosystem and SoftBank\u2019s AITRAS Orchestrator. These years are critical for ensuring that the "carrier-grade" requirements\u201499.999% reliability\u2014are maintained when RAN functions are virtualized on AI-ready hardware.<\/p>\n The planned 2027 trial marks the transition to "real-world" deployment. Unlike previous tests conducted in controlled environments, this trial will involve actual SoftBank subscribers and live data traffic, providing the definitive data set needed for full-scale commercial rollout toward the end of the decade.<\/p>\n The technical backbone of this trial relies on a significant leap in computing power and networking bandwidth. The shift from Purpose-Built RAN to AI-RAN requires hardware that can handle "parallel processing"\u2014a task at which GPUs excel, but traditional CPUs struggle.<\/p>\n The trial will utilize a high-performance hardware configuration designed for the network edge:<\/p>\n A critical component of the trial is the synergy between Nokia MantaRay<\/strong> (Nokia\u2019s network management and optimization suite) and the AITRAS Orchestrator<\/strong>. <\/p>\n In a traditional network, radio resources are managed in a silo. In the AI-RAN model, the Orchestrator must decide in real-time how much power and memory to give to the "Radio" (ensuring your phone call doesn’t drop) versus the "AI" (ensuring a nearby autonomous delivery robot can navigate a sidewalk). Preliminary data suggests that by sharing infrastructure, operators can improve hardware utilization rates from the typical 30\u201340% in traditional RAN to over 80% in an AI-RAN environment.<\/p>\n The leadership of both Nokia and SoftBank view this agreement as a defining moment for the telecommunications industry, positioning the network not just as a "pipe" for data, but as a distributed computer.<\/p>\n Pallavi Mahajan, Chief Technology and AI Officer at Nokia<\/strong>, emphasized the transformative nature of the collaboration:<\/p>\n \u201cOur agreement with SoftBank marks an important step toward the future of AI-native networks. By bringing AI and RAN workloads onto common, programmable infrastructure, AI-RAN has the potential to unlock new levels of spectral efficiency while providing the distributed intelligence needed to support physical AI. As the boundaries between network and compute continue to blur, this distributed infrastructure can become a powerful platform for both connectivity and AI.\u201d<\/p>\n<\/blockquote>\n Mahajan\u2019s comments highlight a strategic pivot for Nokia: moving from a vendor of telecommunications equipment to a provider of "AI-native" software that can run on standardized, accelerated computing platforms.<\/p>\n Ryuji Wakikawa, Senior Vice President and CTO of SoftBank Corp.<\/strong>, focused on the transition from theory to practice:<\/p>\n \u201cOur planned field trial with Nokia will mark a significant milestone in AI-RAN\u2019s transition from concept to execution. We will integrate Nokia\u2019s AI-native anyRAN into our AITRAS platform as its RAN component, with the AITRAS Orchestrator working in coordination with Nokia\u2019s RAN and MantaRay. We are excited to begin this next phase of validation with Nokia, paving the way for a new era of networks that deliver both connectivity and AI services.\u201d<\/p>\n<\/blockquote>\n Wakikawa\u2019s statement underscores SoftBank\u2019s ambition to be more than a mobile carrier; the company aims to become the infrastructure provider for the "AI Revolution" in Japan.<\/p>\n The implications of the Nokia-SoftBank trial extend far beyond the borders of Japan. If successful, this trial will provide the blueprint for the global rollout of 6G and the future of edge computing.<\/p>\n One of the most significant implications of AI-RAN is the support for "Physical AI"\u2014autonomous systems that interact with the real world. This includes self-driving vehicles, industrial robots, and automated logistics. These technologies require massive amounts of AI processing power with near-zero latency. By hosting AI workloads at the base station (the network edge), AI-RAN eliminates the delay caused by sending data to a centralized cloud data center.<\/p>\n For decades, mobile operators have struggled with "monetizing the pipe." While data usage has exploded, revenues have often remained flat. AI-RAN offers a new business model: AI-as-a-Service (AIaaS)<\/strong>. Operators can rent out the excess "compute" on their cell towers to third-party developers, local businesses, or municipal governments to run AI applications, effectively turning every cell site into a mini-data center.<\/p>\n The telecommunications industry is under pressure to reduce its carbon footprint. Traditional RAN hardware is often underutilized but consumes constant power. AI-RAN allows for "dynamic scaling," where energy is consumed only when needed by either the network or the AI applications. Furthermore, AI algorithms can optimize radio beamforming and interference management, allowing more users to share the same spectrum, thus delaying the need for expensive new frequency acquisitions.<\/p>\n While 5G was about "Enhanced Mobile Broadband," 6G is expected to be about "Integrated Sensing and Communication." AI-RAN is the foundational technology for 6G. The 2027 trial will provide the first large-scale data on how AI-native architectures handle the complexities of future 6G waveforms and the massive device density expected in the 2030s.<\/p>\n The agreement between Nokia and SoftBank represents a high-stakes bet on the convergence of silicon and signals. By 2027, the streets of Japan will serve as a laboratory for a new kind of infrastructure\u2014one where the network doesn’t just transport information, but thinks about it. As the trial progresses, the data gathered will likely dictate the technical standards and commercial strategies of the global telecommunications industry for the next decade, marking the true beginning of the AI-native era.<\/p>\n","protected":false},"excerpt":{"rendered":" The global telecommunications landscape is standing on the precipice of a fundamental architectural shift. As the industry moves beyond the initial deployment phases of 5G…<\/p>\n","protected":false},"author":1,"featured_media":2554,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[90],"tags":[698,80,579,584,696,242,999,91,2361,92],"class_list":["post-2557","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-telecommunications","tag-commercial","tag-connectivity","tag-convergence","tag-intelligence","tag-nokia","tag-softbank","tag-stage","tag-telecom","tag-trials","tag-voice"],"_links":{"self":[{"href":"https:\/\/voicecabling.com\/index.php?rest_route=\/wp\/v2\/posts\/2557","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/voicecabling.com\/index.php?rest_route=\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/voicecabling.com\/index.php?rest_route=\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/voicecabling.com\/index.php?rest_route=\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"https:\/\/voicecabling.com\/index.php?rest_route=%2Fwp%2Fv2%2Fcomments&post=2557"}],"version-history":[{"count":0,"href":"https:\/\/voicecabling.com\/index.php?rest_route=\/wp\/v2\/posts\/2557\/revisions"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/voicecabling.com\/index.php?rest_route=\/wp\/v2\/media\/2554"}],"wp:attachment":[{"href":"https:\/\/voicecabling.com\/index.php?rest_route=%2Fwp%2Fv2%2Fmedia&parent=2557"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/voicecabling.com\/index.php?rest_route=%2Fwp%2Fv2%2Fcategories&post=2557"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/voicecabling.com\/index.php?rest_route=%2Fwp%2Fv2%2Ftags&post=2557"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}Core Objectives<\/h3>\n
\n
Chronology: The Road to the 2027 Milestone<\/h2>\n
2022\u20132023: The Conceptual Phase<\/h3>\n
2024: The Year of the AI-RAN Alliance<\/h3>\n
2025\u20132026: Laboratory Refinement and Integration<\/h3>\n
2027: Commercial Field Trial<\/h3>\n
Supporting Data: The Architecture of AI-Native Infrastructure<\/h2>\n
The NVIDIA Hardware Suite<\/h3>\n
\n
Software Integration and Management<\/h3>\n
Official Responses: Insights from Industry Leaders<\/h2>\n
\n
\n
Implications: Defining the AI-Native 6G Era<\/h2>\n
1. The Birth of "Physical AI"<\/h3>\n
2. New Revenue Streams for Operators<\/h3>\n
3. Sustainability and Spectral Efficiency<\/h3>\n
4. The Bridge to 6G<\/h3>\n
Conclusion<\/h2>\n