Domain-specific data based time series foundation model help, specifically in wireless networking.</p>\n","updatedAt":"2026-06-11T17:46:50.019Z","author":{"_id":"65eb7d4e8a38d66e51aacdc3","avatarUrl":"/avatars/7f643bd22fc01077c07a7ef372dd592f.svg","fullname":"Swadhin Pradhan","name":"swadhinp","type":"user","isPro":false,"isHf":false,"isHfAdmin":false,"isMod":false,"isUserFollowing":false}},"numEdits":0,"identifiedLanguage":{"language":"en","probability":0.9200960397720337},"editors":["swadhinp"],"editorAvatarUrls":["/avatars/7f643bd22fc01077c07a7ef372dd592f.svg"],"reactions":[],"isReport":false}}],"primaryEmailConfirmed":false,"paper":{"id":"2606.11553","authors":[{"_id":"6a2af42e4957fcdd3aac0460","name":"Swadhin Pradhan","hidden":false},{"_id":"6a2af42e4957fcdd3aac0461","name":"Niloo Bahadori","hidden":false},{"_id":"6a2af42e4957fcdd3aac0462","name":"Peiman Amini","hidden":false}],"publishedAt":"2026-06-10T01:23:24.000Z","submittedOnDailyAt":"2026-06-11T00:00:00.000Z","title":"APEX: A Network-Native Time-Series Foundation Model for Forecasting and Anomaly Detection for Wireless Edge Operations","submittedOnDailyBy":{"_id":"65eb7d4e8a38d66e51aacdc3","avatarUrl":"/avatars/7f643bd22fc01077c07a7ef372dd592f.svg","isPro":false,"fullname":"Swadhin Pradhan","user":"swadhinp","type":"user","name":"swadhinp"},"summary":"Generic time-series foundation models transfer poorly to wireless network telemetry whose signals are bursty, zero-inflated, and coupled across protocol layers. We present APEX, a network-native, decoder-only transformer for forecasting enterprise AP telemetry, and evaluate it on DHCP degradation as a representative network task. APEX is pre-trained on 10-channel multivariate telemetry from ~4,500 production wireless networks (~100K AP time series, 34 metrics per AP), and is available as APEX-Large (269M, cloud) and APEX-Edge (10.5M, edge). On a 192-step (4-day) DHCP degradation benchmark, APEX-Large reduces MAE by 18% over the strongest foundation-model baseline (Toto) and 38% over SARIMA, with anomaly-detection F1 = 0.93, while APEX-Edge enables sub-second, privacy-preserving inference on AP-class edge hardware. These results suggest network-native pre-training is a practical foundation for proactive wireless operations.","upvotes":1,"discussionId":"6a2af42e4957fcdd3aac0463","ai_summary":"Network-native transformer model APEX demonstrates superior forecasting performance for wireless network telemetry compared to existing foundation models and traditional methods.","ai_keywords":["time-series foundation models","wireless network telemetry","bursty signals","zero-inflated signals","protocol layers","decoder-only transformer","DHCP degradation","multivariate telemetry","pre-trained models","anomaly detection","privacy-preserving inference"],"ai_summary_model":"Qwen/Qwen2.5-Coder-32B-Instruct","organization":{"_id":"6723ed3ab22109dd6c0711d4","name":"ciscoai","fullname":"Cisco","avatar":"https://cdn-avatars.huggingface.co/v1/production/uploads/644f49b2030210812f478a23/hj_fxwQxsfMOeuJIRSx8w.png"}},"canReadDatabase":false,"canManagePapers":false,"canSubmit":false,"hasHfLevelAccess":false,"upvoted":false,"upvoters":[{"_id":"65eb7d4e8a38d66e51aacdc3","avatarUrl":"/avatars/7f643bd22fc01077c07a7ef372dd592f.svg","isPro":false,"fullname":"Swadhin Pradhan","user":"swadhinp","type":"user"}],"acceptLanguages":["en"],"dailyPaperRank":0,"organization":{"_id":"6723ed3ab22109dd6c0711d4","name":"ciscoai","fullname":"Cisco","avatar":"https://cdn-avatars.huggingface.co/v1/production/uploads/644f49b2030210812f478a23/hj_fxwQxsfMOeuJIRSx8w.png"},"markdownContentUrl":"https://huggingface.co/buckets/huggingchat/papers-content/resolve/2606/2606.11553.md"}">
APEX: A Network-Native Time-Series Foundation Model for Forecasting and Anomaly Detection for Wireless Edge Operations
Abstract
Network-native transformer model APEX demonstrates superior forecasting performance for wireless network telemetry compared to existing foundation models and traditional methods.
Generic time-series foundation models transfer poorly to wireless network telemetry whose signals are bursty, zero-inflated, and coupled across protocol layers. We present APEX, a network-native, decoder-only transformer for forecasting enterprise AP telemetry, and evaluate it on DHCP degradation as a representative network task. APEX is pre-trained on 10-channel multivariate telemetry from ~4,500 production wireless networks (~100K AP time series, 34 metrics per AP), and is available as APEX-Large (269M, cloud) and APEX-Edge (10.5M, edge). On a 192-step (4-day) DHCP degradation benchmark, APEX-Large reduces MAE by 18% over the strongest foundation-model baseline (Toto) and 38% over SARIMA, with anomaly-detection F1 = 0.93, while APEX-Edge enables sub-second, privacy-preserving inference on AP-class edge hardware. These results suggest network-native pre-training is a practical foundation for proactive wireless operations.
Community
Domain-specific data based time series foundation model help, specifically in wireless networking.
Upload images, audio, and videos by dragging in the text input, pasting, or clicking here.
Tap or paste here to upload images
Cite arxiv.org/abs/2606.11553 in a model README.md to link it from this page.
Cite arxiv.org/abs/2606.11553 in a dataset README.md to link it from this page.
Cite arxiv.org/abs/2606.11553 in a Space README.md to link it from this page.
Discussion (0)
Sign in to join the discussion. Free account, 30 seconds — email code or GitHub.
Sign in →No comments yet. Sign in and be the first to say something.