{"id":568,"date":"2025-11-18T11:15:20","date_gmt":"2025-11-18T16:15:20","guid":{"rendered":"https:\/\/blogs.mathworks.com\/semiconductors\/?p=568"},"modified":"2025-11-21T09:44:16","modified_gmt":"2025-11-21T14:44:16","slug":"ai-based-receivers-transforming-the-physical-layer-for-6g","status":"publish","type":"post","link":"https:\/\/blogs.mathworks.com\/semiconductors\/2025\/11\/18\/ai-based-receivers-transforming-the-physical-layer-for-6g\/","title":{"rendered":"AI-Based Receivers: Transforming the Physical Layer for 6G"},"content":{"rendered":"<p>As 6G research advances, artificial intelligence is reshaping the physical layer of wireless systems. MathWorks provides tools that facilitate both MATLAB-based AI model development, and integration with other machine learning frameworks like PyTorch\u2122. This gives engineers flexibility without compromising on performance or validation.<\/p>\n<p><div id=\"attachment_570\" style=\"width: 612px\" class=\"wp-caption aligncenter\"><img aria-describedby=\"caption-attachment-570\" decoding=\"async\" loading=\"lazy\" class=\"wp-image-570 size-full\" src=\"http:\/\/blogs.mathworks.com\/semiconductors\/files\/2025\/11\/AInativeReceiver.png\" alt=\"\" width=\"602\" height=\"156\" \/><p id=\"caption-attachment-570\" class=\"wp-caption-text\">AI-native fully convolutional receiver replacing the traditional channel estimation, equalization and symbol demodulation to compensate for the effects of the propagation channel<\/p><\/div><\/p>\n<pre>Interested to learn more? Explore the examples:\r\n<a href=\"https:\/\/www.mathworks.com\/help\/5g\/ug\/ai-native-fully-convolutional-receiver.html\">AI-Native Fully Convolutional Receiver<\/a>\r\n<a href=\"https:\/\/www.mathworks.com\/help\/5g\/ug\/verify-performance-of-6g-ai-native-receiver-using-pytorch.html\">Verify the Performance of 6G AI-Native Receiver Using MATLAB and PyTorch Coexecution<\/a><\/pre>\n<p><!--more--><\/p>\n<h2><strong>Two Paths, One Platform<\/strong><\/h2>\n<p>MathWorks showcases this flexibility through two examples:<\/p>\n<ul>\n<li><a href=\"https:\/\/www.mathworks.com\/help\/5g\/ug\/ai-native-fully-convolutional-receiver.html\"><strong>AI-Native Fully Convolutional Receiver<\/strong><\/a>: Built entirely in MATLAB, this example shows how to train a CNN to replace traditional receiver blocks. It\u2019s optimized for challenging conditions like high Doppler and sparse pilots.<\/li>\n<li><a href=\"https:\/\/www.mathworks.com\/help\/5g\/ug\/verify-performance-of-6g-ai-native-receiver-using-pytorch.html\"><strong>Verify Performance of 6G AI-Native Receiver using MATLAB and PyTorch Coexecution<\/strong><\/a>: For users with PyTorch-trained models, this example shows how to measure the performance of a PyTorch trained AI-native receiver with standard 5G waveforms, enabling performance benchmarking under realistic channel conditions.<\/li>\n<\/ul>\n<p><div id=\"attachment_571\" style=\"width: 469px\" class=\"wp-caption aligncenter\"><img aria-describedby=\"caption-attachment-571\" decoding=\"async\" loading=\"lazy\" class=\"wp-image-571 size-full\" src=\"http:\/\/blogs.mathworks.com\/semiconductors\/files\/2025\/11\/AInativeReceiverExample.png\" alt=\"\" width=\"459\" height=\"280\" \/><p id=\"caption-attachment-571\" class=\"wp-caption-text\">AI-native receiver examples from 6G Exploration Library for 5G Toolbox<\/p><\/div><\/p>\n<p><div id=\"attachment_572\" style=\"width: 510px\" class=\"wp-caption aligncenter\"><img aria-describedby=\"caption-attachment-572\" decoding=\"async\" loading=\"lazy\" class=\"wp-image-572 size-full\" src=\"http:\/\/blogs.mathworks.com\/semiconductors\/files\/2025\/11\/AInativeReceiverPerformance.png\" alt=\"\" width=\"500\" height=\"257\" \/><p id=\"caption-attachment-572\" class=\"wp-caption-text\">Performance of fully convolutional receiver vs. traditional receiver for a high Doppler case<\/p><\/div><\/p>\n<h2><strong>MATLAB: a Unified Environment for AI for Wireless Simulation<\/strong><\/h2>\n<p>Whether you build your model in MATLAB or in PyTorch, MathWorks tools provide a consistent simulation environment:<\/p>\n<ul>\n<li>Generate synthetic data for learning and inference<\/li>\n<li>Capture real world data for AI training and validation using high end USRP radios and <a href=\"https:\/\/www.mathworks.com\/products\/wireless-testbench.html\">Wireless Testbench<\/a><\/li>\n<li>Accelerate simulations using parallel processing and GPUs<\/li>\n<li><a href=\"https:\/\/www.mathworks.com\/solutions\/deep-learning\/embedded-ai.html\">Deploy AI networks<\/a> across embedded platforms<\/li>\n<\/ul>\n<p><div id=\"attachment_573\" style=\"width: 762px\" class=\"wp-caption aligncenter\"><img aria-describedby=\"caption-attachment-573\" decoding=\"async\" loading=\"lazy\" class=\"wp-image-573 size-full\" src=\"http:\/\/blogs.mathworks.com\/semiconductors\/files\/2025\/11\/AIsimulationUnifiedEnvironment.png\" alt=\"\" width=\"752\" height=\"161\" \/><p id=\"caption-attachment-573\" class=\"wp-caption-text\">With MATLAB and Simulink, you can design, simulate, test, verify, and deploy AI algorithms that enhance the performance and functionality of complex embedded systems<\/p><\/div><\/p>\n<h2><strong>Supporting Diverse Workflows for 6G Innovation<\/strong><\/h2>\n<p>From prototyping in MATLAB to validating external models, MathWorks provides a simulation framework that accelerates your path to 6G innovation.<\/p>\n<ul>\n<li>Use MATLAB\u2019s Deep Learning Toolbox to build and train models entirely within the MATLAB environment<\/li>\n<li>Bring your own models from PyTorch or TensorFlow\u2122 and run inference directly within MATLAB simulations using co-execution workflows<\/li>\n<li>Explore more <a href=\"https:\/\/www.mathworks.com\/help\/5g\/examples.html?category=ai-for-6g&amp;exampleproduct=all&amp;newonly=&amp;s_tid=CRUX_lftnav\">AI for 6G examples<\/a> and <a href=\"https:\/\/www.mathworks.com\/help\/comm\/examples.html?category=ai-for-wireless&amp;exampleproduct=all&amp;newonly=&amp;s_tid=CRUX_lftnav\">AI for wireless examples<\/a> offering a growing library of examples covering AI-based beamforming, channel estimation, signal classification, and more<\/li>\n<\/ul>\n<p><div id=\"attachment_574\" style=\"width: 612px\" class=\"wp-caption aligncenter\"><img aria-describedby=\"caption-attachment-574\" decoding=\"async\" loading=\"lazy\" class=\"wp-image-574 size-full\" src=\"http:\/\/blogs.mathworks.com\/semiconductors\/files\/2025\/11\/CSIwithAutoencoderExample.png\" alt=\"\" width=\"602\" height=\"357\" \/><p id=\"caption-attachment-574\" class=\"wp-caption-text\">Image from the CSI Feedback with Autoencoders example, where we use \u00a0neural networks to compress and recover channel state information between the UE and the gNB. The figure shows the preprocessing of the channel state information in the frequency domain before compressing using autoencoders<\/p><\/div><\/p>\n<h2>Conclusion<\/h2>\n<p>AI is redefining the physical layer design. MathWorks provides a unified simulation framework to build, test, and validate AI-driven algorithms\u2014whether you create models in MATLAB or bring in PyTorch for co-execution. With AI shaping the future of connectivity, the question is: how will you use these tools to accelerate your 6G journey?<\/p>\n<pre>Interested to learn more? Explore the examples:\r\n<a href=\"https:\/\/www.mathworks.com\/help\/5g\/ug\/ai-native-fully-convolutional-receiver.html\">AI-Native Fully Convolutional Receiver<\/a>\r\n<a href=\"https:\/\/www.mathworks.com\/help\/5g\/ug\/verify-performance-of-6g-ai-native-receiver-using-pytorch.html\">Verify the Performance of 6G AI-Native Receiver Using MATLAB and PyTorch Coexecution<\/a><\/pre>\n<p><!--more--><\/p>\n","protected":false},"excerpt":{"rendered":"<div class=\"overview-image\"><img decoding=\"async\"  class=\"img-responsive\" src=\"http:\/\/blogs.mathworks.com\/semiconductors\/files\/2025\/11\/AInativeReceiver.png\" onError=\"this.style.display ='none';\" \/><\/div>\n<p>As 6G research advances, artificial intelligence is reshaping the physical layer of wireless systems. MathWorks provides tools that facilitate both MATLAB-based AI model development, and integration&#8230; <a class=\"read-more\" href=\"https:\/\/blogs.mathworks.com\/semiconductors\/2025\/11\/18\/ai-based-receivers-transforming-the-physical-layer-for-6g\/\">read more >><\/a><\/p>\n","protected":false},"author":221,"featured_media":0,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":[],"categories":[2],"tags":[],"_links":{"self":[{"href":"https:\/\/blogs.mathworks.com\/semiconductors\/wp-json\/wp\/v2\/posts\/568"}],"collection":[{"href":"https:\/\/blogs.mathworks.com\/semiconductors\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/blogs.mathworks.com\/semiconductors\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/blogs.mathworks.com\/semiconductors\/wp-json\/wp\/v2\/users\/221"}],"replies":[{"embeddable":true,"href":"https:\/\/blogs.mathworks.com\/semiconductors\/wp-json\/wp\/v2\/comments?post=568"}],"version-history":[{"count":21,"href":"https:\/\/blogs.mathworks.com\/semiconductors\/wp-json\/wp\/v2\/posts\/568\/revisions"}],"predecessor-version":[{"id":597,"href":"https:\/\/blogs.mathworks.com\/semiconductors\/wp-json\/wp\/v2\/posts\/568\/revisions\/597"}],"wp:attachment":[{"href":"https:\/\/blogs.mathworks.com\/semiconductors\/wp-json\/wp\/v2\/media?parent=568"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/blogs.mathworks.com\/semiconductors\/wp-json\/wp\/v2\/categories?post=568"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/blogs.mathworks.com\/semiconductors\/wp-json\/wp\/v2\/tags?post=568"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}