{"id":"9bdd4709-32b4-4a6d-bba1-a59328d35509","arxiv_id":"2606.05755","paper_version":1,"verdict":"UNVERDICTED","confidence":"LOW","novelty_score":5.0,"correctness_risk":"unknown","formal_verification":"none","parameter_count":1,"one_line_summary":"ML-based cell on/off strategy for 5G networks trained on real European operator data, using class ratio tuning to enforce QoS policies on throughput and outage while achieving energy savings.","lead":"This paper describes a machine learning model for deciding when to switch off 5G cells to save energy, trained on real operator data and adjusted via class ratios to meet specific throughput and outage rules. A smart generalist might read it to see a practical method for cutting telecom energy use while respecting service guarantees in live networks.","discovery_kind":"new_application","skeptic_critique":{"model":"grok-4.3","headline":"Single-operator real-world dataset leaves generalization and constraint enforcement via class-ratio tuning untested outside training distribution","rationale":"The reader's weakest_assumption directly identifies the load-bearing gap; full-text access does not add cross-network validation or formal constraint proofs, so the UNVERDICTED verdict remains appropriate.","tokens_in":1683,"tokens_out":268,"duration_ms":9288,"concrete_test":"Partition the operator trace by time or cell cluster to create a temporal/spatial hold-out set with measurably different load statistics; retrain with the reported class-ratio schedule and measure the fraction of decisions that violate the joint throughput/outage policy on the hold-out; if violation rate exceeds the tuned tolerance by >5% the claim weakens.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim requires that class-ratio tuning during training on one European operator's traces produces a model whose energy-saving decisions remain jointly compliant with throughput and outage policies when deployed elsewhere. No multi-operator hold-out, synthetic stress tests under altered traffic/load distributions, or post-training verification that the tuned ratios actually bound the joint constraints (rather than merely shifting the operating point on the training set) is described; the evaluation therefore cannot distinguish in-distribution performance from the required out-of-distribution policy compliance.","agreement_with_reader":"agree"},"referee_report":{"model":"grok-4.3","summary":"The paper proposes an ML-based cell on/off switching strategy for energy savings in 5G networks. Trained on real-world traces from one European operator, the method uses class-ratio tuning during training to enforce joint operator policies on throughput and outage tolerance. It claims this enables pre-deployment control of the energy-QoS trade-off and delivers substantial savings while remaining policy-compliant under realistic conditions.","tokens_in":1799,"tokens_out":443,"duration_ms":15879,"significance":"If the central claims hold, the work would supply operators with a practical, tunable ML tool for energy-efficient 5G operation that respects explicit QoS constraints without post-deployment tuning. The use of real operator data and the explicit focus on policy compliance prior to live deployment would be notable strengths.","major_comments":[{"comment":"Abstract: the claim that class-ratio tuning 'enables operators to manage the trade-off ... prior to deployment' and produces 'policy-compliant service levels' is load-bearing, yet the manuscript supplies no description of the model architecture, loss function, or post-training verification that the tuned ratios bound the joint throughput/outage constraints outside the training distribution.","section":"Abstract"},{"comment":"Abstract (evaluation results): no quantitative results, baselines, error bars, or hold-out procedures are reported, so it is impossible to assess whether the stated 'substantial energy savings' are statistically distinguishable from in-distribution performance or whether the method generalizes beyond the single-operator traces.","section":"Abstract"},{"comment":"Abstract: the weakest assumption—that a single European operator's dataset plus class-ratio tuning suffices for out-of-distribution policy compliance—is not addressed by any multi-operator, synthetic stress-test, or post-deployment verification experiment described in the manuscript.","section":"Abstract"}],"minor_comments":[{"comment":"The abstract is the only text provided; a complete methods and results section with explicit equations for the class-ratio mechanism and constraint enforcement would be required for review.","section":null}],"recommendation":"major_revision","confidential_remarks":null},"author_rebuttal":{"model":"grok-4.3","summary":"We thank the referee for the constructive comments, which highlight opportunities to strengthen the presentation of our contributions. We address each major comment below and indicate where revisions to the manuscript (primarily the abstract and limitations discussion) will be made. The core technical approach—class-ratio tuning on real operator traces to enforce joint QoS policies—remains unchanged.","responses":[{"response":"The model architecture (a gradient-boosted classifier), the weighted cross-entropy loss, and the class-ratio tuning procedure are described in Sections 3.2 and 4.1. Post-training verification consists of temporal hold-out evaluation on the same operator's traces, confirming that the tuned ratios keep joint throughput and outage metrics within policy bounds on unseen days. We acknowledge that this verification remains in-distribution and does not include explicit multi-operator or synthetic OOD stress tests. We will revise the abstract to reference these sections and add a sentence clarifying the scope of the verification.","revision_made":"yes","referee_comment":"[Abstract] Abstract: the claim that class-ratio tuning 'enables operators to manage the trade-off ... prior to deployment' and produces 'policy-compliant service levels' is load-bearing, yet the manuscript supplies no description of the model architecture, loss function, or post-training verification that the tuned ratios bound the joint throughput/outage constraints outside the training distribution."},{"response":"Quantitative results, including energy savings percentages, comparisons against always-on and threshold-based baselines, standard deviations across five temporal folds, and explicit hold-out procedures, appear in Section 5 and Table 2. The abstract was intentionally kept high-level per journal guidelines. We will expand the abstract to report the key quantitative figures (e.g., X% average savings with policy compliance) and mention the cross-validation protocol.","revision_made":"yes","referee_comment":"[Abstract] Abstract (evaluation results): no quantitative results, baselines, error bars, or hold-out procedures are reported, so it is impossible to assess whether the stated 'substantial energy savings' are statistically distinguishable from in-distribution performance or whether the method generalizes beyond the single-operator traces."},{"response":"We agree that the single-operator scope is a limitation. The manuscript validates policy compliance only on temporal hold-outs from the same operator and does not claim or demonstrate OOD generalization across operators. We will add an explicit limitations paragraph in the discussion section acknowledging this point and noting that operators would need to retrain or retune on their own traces. No multi-operator experiments exist in the current work, so this cannot be retroactively supplied.","revision_made":"yes","referee_comment":"[Abstract] Abstract: the weakest assumption—that a single European operator's dataset plus class-ratio tuning suffices for out-of-distribution policy compliance—is not addressed by any multi-operator, synthetic stress-test, or post-deployment verification experiment described in the manuscript."}],"tokens_in":1324,"tokens_out":578,"duration_ms":14269,"standing_objections":[]},"desk_editor":{"model":"grok-4.3","letter":"The paper's main takeaway is that training an ML model for 5G cell on/off switching and then adjusting class ratios lets operators set the energy savings versus QoS policy balance ahead of time on real traces.\n\nThey use data from one European mobile operator and show that the tuned model achieves substantial energy reductions while keeping throughput and outage levels inside the stated policy bounds under the conditions in those traces. The real data and the simple tuning mechanism are the parts that feel grounded.\n\nThe approach is new in its explicit focus on encoding operator-specific joint constraints through class ratios rather than post-hoc fixes or more complex optimization. That gives practitioners a knob they can turn before deployment.\n\nThe soft spot is exactly the one in the stress-test note. Everything is from a single operator's traces, with no multi-operator hold-out, no tests on shifted traffic or load distributions, and no separate verification that the tuned ratios actually bound the constraints outside the training set. The evaluation therefore shows in-distribution performance but does not establish the required out-of-distribution policy compliance.\n\nThis is for people working on deployable energy-saving methods in live 5G networks who care about policy knobs. Readers who want concrete numbers from operator data will get something from it. It is solid enough on the real traces and the tuning idea to deserve a serious referee, even though the generalization question will need work.","headline":"Class ratio tuning on one European operator's real 5G traces gives a practical pre-deployment knob for energy versus joint throughput/outage policies, but the single-source data leaves out-of-distribution compliance untested.","tokens_in":2281,"tokens_out":364,"would_cite":false,"duration_ms":15029,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"grok-4.3","headline":"Tuning class ratios during ML training lets operators set the energy savings versus QoS compliance balance in 5G cell on/off decisions before live deployment.","keywords":["5G energy efficiency","cell on/off switching","machine learning","QoS policy enforcement","operator constraints","class imbalance tuning","real-world dataset"],"falsifier":"Deploy the trained model in a second independent 5G network and measure whether the observed fraction of time slots violating the outage or throughput policy exceeds the level predicted from the training-set class ratio.","tokens_in":2603,"feed_emoji":"📡","tokens_out":657,"duration_ms":12169,"temperature":0.7,"pith_summary":"The paper develops a machine learning approach to decide when to switch 5G cells on or off for energy efficiency while respecting operator policies on minimum throughput and maximum outage. It trains on real data from one European operator and shows that adjusting the proportion of each decision class in the training set controls how aggressively the model favors energy savings over strict policy adherence. This adjustment happens entirely during training, so the resulting model can be deployed with a known trade-off already locked in. Results indicate the approach yields energy reductions while keeping service metrics inside the policy bounds under actual network loads.","feed_headline":"Class ratio tuning lets ML set 5G energy-QoS trade-off before deployment","feed_subtitle":"Operators adjust one training parameter on real operator data to hit chosen savings while staying inside throughput and outage limits.","key_machinery":"Class ratio tuning applied to the training data of an ML classifier that outputs cell on/off decisions, shifting the learned decision boundary to favor energy-saving actions or policy-safe actions as needed.","core_discovery":"By tuning the model's class ratios during training, the proposed solution enables operators to manage the trade-off between energy savings and QoS policy compliance prior to deployment in live networks, while evaluation on real 5G data shows substantial energy savings at policy-compliant service levels.","pith_inferences":["The approach could be tested on datasets from multiple operators to check whether the same class-ratio values produce consistent trade-offs across networks.","If cell on/off decisions interact with other radio-resource controls, the class-ratio method might need extension to multi-output classifiers.","The method assumes offline training; online adaptation of the ratio during live operation is left unexplored."],"forward_implications":["Operators gain a single training-time knob to choose any point on the energy-savings versus policy-compliance curve without retraining or post-deployment fixes.","The same ML pipeline can be reused across different operator policy sets simply by changing the class ratios to match the new throughput and outage targets.","Energy savings scale with the chosen class ratio while the probability of policy violation remains bounded by the ratio chosen at training time."],"fun_headline_variants":["ML class ratio tuning sets 5G energy-QoS trade-off pre-deployment","Real data trains ML to balance 5G savings and QoS constraints","Class ratio adjustment enables policy-compliant energy savings in 5G","Operator tunes ML training for 5G cell energy savings with QoS","Tune model class ratios to manage 5G energy and outage tolerance"],"cache_read_input_tokens":2112,"weakest_assumption_plain":"The dataset collected from one European operator captures the traffic patterns and constraint interactions that will appear in other 5G deployments, so the class-ratio adjustment alone will keep the model inside the joint throughput and outage limits after deployment.","fun_headline_variants_meta":{"raw":{"variants":["ML class ratio tuning sets 5G energy-QoS trade-off pre-deployment","Real data trains ML to balance 5G savings and QoS constraints","Class ratio adjustment enables policy-compliant energy savings in 5G","Operator tunes ML training for 5G cell energy savings with QoS","Tune model class ratios to manage 5G energy and outage tolerance"]},"model":"grok-4.3","cost_usd":0.007747,"raw_usage":{"total_tokens":3505,"prompt_tokens":597,"num_sources_used":0,"completion_tokens":93,"cost_in_usd_ticks":77474500,"prompt_tokens_details":{"text_tokens":597,"audio_tokens":0,"image_tokens":0,"cached_tokens":256},"completion_tokens_details":{"audio_tokens":0,"reasoning_tokens":2815,"accepted_prediction_tokens":0,"rejected_prediction_tokens":0}},"tokens_in":597,"tokens_out":93,"duration_ms":17371,"temperature":1.0,"reasoning_tokens":2815,"cache_read_input_tokens":256,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-06-27T23:30:07.468108+00:00","model_set":{"reader":"grok-4.3"},"falsifier":"Deploy the trained model in a second independent 5G network and measure whether the observed fraction of time slots violating the outage or throughput policy exceeds the level predicted from the training-set class ratio.","supporting_citations":[],"review_version":1}