{"slug": "ericsson-runs-ai-native-link-adaptation-scheduler-trial-with-softbank", "title": "Ericsson runs AI-native link adaptation scheduler trial with SoftBank", "summary": "Ericsson and SoftBank completed Japan's first trial of Ericsson's AI-native Scheduler for Link Adaptation on SoftBank's commercial 5G network, achieving up to a 25% improvement in spectral efficiency and up to a 50% increase in downlink user throughput, with a 10% improvement in each parameter across all trial locations. The trial validates AI's role in enhancing RAN performance as the companies collaborate toward 5G-Advanced and 6G.", "body_md": "Japanese telecom operator SoftBank and Ericsson conducted the first trial in that country of an AI-native link adaptation scheduler that showed a significant improvement in radio access network (RAN) performance and efficiency.\n\nThe trial used Ericsson’s AI-native Scheduler for Link Adaptation system, which is part of the vendor’s AI in RAN software. The system was injected into SoftBank’s commercial 5G network, running on that network’s baseband equipment to manage network resources in real time.\n\nThe system ingests network performance parameters and uses an AI algorithm to adjust network resources to deal with changing radio environments. This includes the ability to optimize link parameters for users located at cell edges or in areas with high interference; stabilize network throughput under adverse radio conditions such as at cell edges and in congested areas; and allow for more traffic within existing frequency bands through improved spectral efficiency.\n\nThe trial showed up to a 25% improvement in spectral efficiency and up to a 50% increase in downlink user throughput compared with legacy systems. More broadly, the companies touted a 10% improvement in each parameter across all locations used in the trial.\n\n“Through this validation, SoftBank and Ericsson confirmed that AI can contribute to the realization of high-performing, secure, and flexible networks as the two companies collaborate toward 5G-Advanced and 6G in the future,” they noted in a statement.\n\nEricsson [conducted a similar trial earlier this year](https://www.sdxcentral.com/news/t-mobile-us-ericsson-ramp-ai-ran-efficiency-efforts/) with T-Mobile US.\n\nThat work integrated an AI-native scheduler and link adaption into “large-scale commercial trials” on live 5G-Advanced network traffic. The software tapped neural networks running on Ericsson “optimized” hardware that can predict changing radio conditions in real time, with the trial showing a 10% improvement in spectral efficiency and 15% increase in downlink speeds compared with “legacy rule-based methods.”\n\nThose stats matched previous technology trial results conducted in more limited conditions, “demonstrating the solution’s adaptability to diverse environments as the two companies jointly broaden the footprint further.”\n\n## Different, but the same\n\nJapan’s NTT DoCoMo recently ran a similar trial with equipment vendor Samsung.\n\nThat test operated AI-RAN optimization technology at the individual-user level rather than applying blanket configurations across an entire cell, automatically applying configurations tailored to real-time wireless conditions and the services being used. Instead of focusing on radio link failure prediction, a customer-experience-driven approach was adopted aimed at supporting uninterrupted video streaming and other latency-heavy services.\n\nThe system used aggregated data from user devices to analyze individual user contexts and movement trends to flag early signs of throughput degradation before service quality declines. According to DoCoMo, the AI element proactively predicted network degradation before performance drops occurred, with compute run on Samsung hardware. In their Japan-based field test, the companies claimed to have reduced the frequency of communication-speed degradation from 13.1% to 7.2% when compared with traditional operations.\n\nDoCoMo and Samsung also developed a data collection process to minimize network congestion during wireless network optimization by selectively gathering only essential information based on specific user issues. Eschewing the traditional bulk collection and processing of wireless environment data, the method was claimed to reduce strain on the network while gathering the critical data needed for the AI-RAN operation.", "url": "https://wpnews.pro/news/ericsson-runs-ai-native-link-adaptation-scheduler-trial-with-softbank", "canonical_source": "https://www.sdxcentral.com/news/ericsson-runs-ai-native-link-adaptation-scheduler-trial-with-softbank/", "published_at": "2026-08-20 15:55:14+00:00", "updated_at": "2026-08-20 16:13:13.690530+00:00", "lang": "en", "topics": ["artificial-intelligence", "ai-products", "ai-infrastructure", "machine-learning"], "entities": ["Ericsson", "SoftBank", "T-Mobile US", "NTT DoCoMo", "Samsung"], "alternates": {"html": "https://wpnews.pro/news/ericsson-runs-ai-native-link-adaptation-scheduler-trial-with-softbank", "markdown": "https://wpnews.pro/news/ericsson-runs-ai-native-link-adaptation-scheduler-trial-with-softbank.md", "text": "https://wpnews.pro/news/ericsson-runs-ai-native-link-adaptation-scheduler-trial-with-softbank.txt", "jsonld": "https://wpnews.pro/news/ericsson-runs-ai-native-link-adaptation-scheduler-trial-with-softbank.jsonld"}}