{"corpus_id":191193864,"paper_sha":"e4d7d557562504cf2bf12f908cb4ed97741c558d","doi":"10.3390/APP9102171","arxiv_id":null,"pmid":null,"pmcid":null,"mag_id":2947150109,"dblp_id":null,"acl_id":null,"title":"Automatic Digital Modulation Classification Based on Curriculum Learning","year":2019,"publication_date":"2019-05-27","venue":"Applied Sciences","journal":{"name":"Applied Sciences","pages":null,"volume":null},"journal_issn":null,"journal_title":null,"publication_types":[],"pubmed_pub_types":null,"s2_fields_of_study":["Computer Science","Engineering"],"reference_count":29,"citation_count":24,"influential_citation_count":1,"is_open_access":true,"arxiv_categories":null,"arxiv_license":null,"arxiv_journal_ref":null,"mesh_headings":null,"chemicals":null,"comments_corrections":null,"source_flags":1,"s2_open_access_pdf_url":"https://www.mdpi.com/2076-3417/9/10/2171/pdf?version=1559099695","s2_open_access_landing_url":"https://www.semanticscholar.org/paper/e4d7d557562504cf2bf12f908cb4ed97741c558d","s2_open_access_license":"CCBY","s2_open_access_status":"GOLD","pmc_open_access_pdf_url":null,"pmc_open_access_landing_url":null,"pmc_open_access_license":null,"pmc_open_access_status":null,"unpaywall_open_access_pdf_url":null,"unpaywall_open_access_landing_url":null,"unpaywall_open_access_license":null,"unpaywall_open_access_status":null,"abstract":"Neural network shows great potential in modulation classification because of its excellent accuracy and achievability but overfitting and memorizing data noise often happen in previous researches on automatic digital modulation classifier. To solve this problem, we utilize two neural networks, namely MentorNet and StudentNet, to construct an automatic modulation classifier, which possesses great performance on the test set with −18–20 dB signal-to-noise ratio (SNR). The MentorNet supervises the training of StudentNet according to curriculum learning, and deals with the overfitting problem in StudentNet. The proposed classifier is verified in several test sets containing additive white Gaussian noise (AWGN), Rayleigh fading, carrier frequency offset and phase offset. Experimental results reveal that the overall accuracy of this classifier for common eleven modulation types was up to 99.3% while the inter-class accuracy could be up to 100%, which was much higher than many other classifiers. Besides, in the presence of interferences, the overall accuracy of this novel classifier still could reach 90% at 10 dB SNR indicting its excellent robustness, which makes it suitable for applications like military electronic warfare.","claims":[{"public_id":"cl_c6a3a41fa0aeb34f3577f28a9e37a0ef","status":"active","text":"A MentorNet-StudentNet curriculum-learning classifier achieves strong performance on test sets at signal-to-noise ratios from −18 to 20 dB.","confidence":0.96,"contributors":[{"id":1,"public_id":"12632b8b5f","public_label":"Anonymous (12632b8b5f)","roles":["extraction"],"url":"https://sah.borca.ai/u/12632b8b5f"}],"url":"https://sah.borca.ai/claims/cl_c6a3a41fa0aeb34f3577f28a9e37a0ef"},{"public_id":"cl_5be67b565554a196221b5e5516b1f765","status":"active","text":"MentorNet supervises StudentNet training and addresses overfitting in StudentNet.","confidence":0.95,"contributors":[{"id":1,"public_id":"12632b8b5f","public_label":"Anonymous (12632b8b5f)","roles":["extraction"],"url":"https://sah.borca.ai/u/12632b8b5f"}],"url":"https://sah.borca.ai/claims/cl_5be67b565554a196221b5e5516b1f765"},{"public_id":"cl_7edb5e4850083b4d089b9b20101b20bd","status":"active","text":"Overall accuracy for eleven common modulation types reached 99.3%, inter-class accuracy reached 100%, and overall accuracy remained 90% at 10 dB SNR under interference.","confidence":0.98,"contributors":[{"id":1,"public_id":"12632b8b5f","public_label":"Anonymous (12632b8b5f)","roles":["extraction"],"url":"https://sah.borca.ai/u/12632b8b5f"}],"url":"https://sah.borca.ai/claims/cl_7edb5e4850083b4d089b9b20101b20bd"},{"public_id":"cl_05a0744b31c878af58fe34f857921b77","status":"active","text":"The classifier was validated on test sets containing additive white Gaussian noise, Rayleigh fading, carrier frequency offset, and phase offset.","confidence":0.94,"contributors":[{"id":1,"public_id":"12632b8b5f","public_label":"Anonymous (12632b8b5f)","roles":["extraction"],"url":"https://sah.borca.ai/u/12632b8b5f"}],"url":"https://sah.borca.ai/claims/cl_05a0744b31c878af58fe34f857921b77"}],"concepts":[{"public_id":"co_09f89e06a2621eec3874624089378d56","status":"active","name":"StudentNet","description":"A neural network trained under MentorNet supervision in the proposed classifier.","types":["neural network","component"],"aliases":[],"contributors":[{"id":1,"public_id":"12632b8b5f","public_label":"Anonymous (12632b8b5f)","roles":["extraction"],"url":"https://sah.borca.ai/u/12632b8b5f"}],"url":"https://sah.borca.ai/concepts/co_09f89e06a2621eec3874624089378d56"},{"public_id":"co_28e96a14d0cc89ab8b588c5ecf7f120d","status":"active","name":"overfitting","description":"A training issue in which a model fits noise or idiosyncrasies of the training data too closely.","types":["training problem"],"aliases":[],"contributors":[{"id":1,"public_id":"12632b8b5f","public_label":"Anonymous (12632b8b5f)","roles":["extraction"],"url":"https://sah.borca.ai/u/12632b8b5f"}],"url":"https://sah.borca.ai/concepts/co_28e96a14d0cc89ab8b588c5ecf7f120d"},{"public_id":"co_37ff067be824d625714db691275c6851","status":"active","name":"phase offset","description":"A constant phase mismatch between transmitted and received signals.","types":["impairment"],"aliases":[],"contributors":[{"id":1,"public_id":"12632b8b5f","public_label":"Anonymous (12632b8b5f)","roles":["extraction"],"url":"https://sah.borca.ai/u/12632b8b5f"}],"url":"https://sah.borca.ai/concepts/co_37ff067be824d625714db691275c6851"},{"public_id":"co_3a0accb02488314099001b668ebf9bb9","status":"active","name":"additive white Gaussian noise","description":"Random Gaussian noise added to a signal with constant spectral density.","types":["noise condition"],"aliases":["AWGN"],"contributors":[{"id":1,"public_id":"12632b8b5f","public_label":"Anonymous (12632b8b5f)","roles":["extraction"],"url":"https://sah.borca.ai/u/12632b8b5f"}],"url":"https://sah.borca.ai/concepts/co_3a0accb02488314099001b668ebf9bb9"},{"public_id":"co_3a8d284e3234c07baa46264f4d2bd3f8","status":"active","name":"eleven modulation types","description":"The set of eleven common digital modulation classes used in the evaluation.","types":["class set"],"aliases":[],"contributors":[{"id":1,"public_id":"12632b8b5f","public_label":"Anonymous (12632b8b5f)","roles":["extraction"],"url":"https://sah.borca.ai/u/12632b8b5f"}],"url":"https://sah.borca.ai/concepts/co_3a8d284e3234c07baa46264f4d2bd3f8"},{"public_id":"co_412fa39072382db73cd3dc29503ef049","status":"active","name":"carrier frequency offset","description":"A mismatch between transmitter and receiver carrier frequencies that shifts the received signal spectrum.","types":["impairment"],"aliases":["CFO"],"contributors":[{"id":1,"public_id":"12632b8b5f","public_label":"Anonymous (12632b8b5f)","roles":["extraction"],"url":"https://sah.borca.ai/u/12632b8b5f"}],"url":"https://sah.borca.ai/concepts/co_412fa39072382db73cd3dc29503ef049"},{"public_id":"co_5fe752e79d3d4874582aecfb2c66665f","status":"active","name":"MentorNet","description":"A neural network used to supervise the training process of another network in the proposed classifier.","types":["neural network","component"],"aliases":[],"contributors":[{"id":1,"public_id":"12632b8b5f","public_label":"Anonymous (12632b8b5f)","roles":["extraction"],"url":"https://sah.borca.ai/u/12632b8b5f"}],"url":"https://sah.borca.ai/concepts/co_5fe752e79d3d4874582aecfb2c66665f"},{"public_id":"co_715b73f566e10d9b90f136f5a07d1d46","status":"active","name":"Rayleigh fading","description":"A wireless channel fading model characterized by Rayleigh-distributed signal amplitudes.","types":["channel condition"],"aliases":[],"contributors":[{"id":1,"public_id":"12632b8b5f","public_label":"Anonymous (12632b8b5f)","roles":["extraction"],"url":"https://sah.borca.ai/u/12632b8b5f"}],"url":"https://sah.borca.ai/concepts/co_715b73f566e10d9b90f136f5a07d1d46"},{"public_id":"co_85293248ab1a02b08947ad0e6c9860ca","status":"active","name":"overall accuracy","description":"The proportion of all test instances classified correctly.","types":["evaluation metric"],"aliases":[],"contributors":[{"id":1,"public_id":"12632b8b5f","public_label":"Anonymous (12632b8b5f)","roles":["extraction"],"url":"https://sah.borca.ai/u/12632b8b5f"}],"url":"https://sah.borca.ai/concepts/co_85293248ab1a02b08947ad0e6c9860ca"},{"public_id":"co_97c056db090f05eacdaab8869a477373","status":"active","name":"inter-class accuracy","description":"A class-wise accuracy measure reported for the modulation classifier.","types":["evaluation metric"],"aliases":[],"contributors":[{"id":1,"public_id":"12632b8b5f","public_label":"Anonymous (12632b8b5f)","roles":["extraction"],"url":"https://sah.borca.ai/u/12632b8b5f"}],"url":"https://sah.borca.ai/concepts/co_97c056db090f05eacdaab8869a477373"},{"public_id":"co_a3a0925760220e5ffdc1d88f699b034b","status":"active","name":"automatic digital modulation classification","description":"The task of identifying a digital modulation scheme from received signal data.","types":["task"],"aliases":["modulation classification","automatic modulation classification"],"contributors":[{"id":1,"public_id":"12632b8b5f","public_label":"Anonymous (12632b8b5f)","roles":["extraction"],"url":"https://sah.borca.ai/u/12632b8b5f"}],"url":"https://sah.borca.ai/concepts/co_a3a0925760220e5ffdc1d88f699b034b"},{"public_id":"co_bd4f7cdbd28f54fc19aafdb4341a4daf","status":"active","name":"curriculum learning","description":"A training strategy that presents examples in a learned or ordered progression of difficulty.","types":["training strategy"],"aliases":[],"contributors":[{"id":1,"public_id":"12632b8b5f","public_label":"Anonymous (12632b8b5f)","roles":["extraction"],"url":"https://sah.borca.ai/u/12632b8b5f"}],"url":"https://sah.borca.ai/concepts/co_bd4f7cdbd28f54fc19aafdb4341a4daf"}],"external_ids":{"DOI":"10.3390/APP9102171","ArXiv":null,"PubMed":null,"PubMedCentral":null,"MAG":2947150109,"DBLP":null,"ACL":null},"open_access":{"is_open_access":true,"pdf_url":"https://www.mdpi.com/2076-3417/9/10/2171/pdf?version=1559099695","landing_url":"https://www.semanticscholar.org/paper/e4d7d557562504cf2bf12f908cb4ed97741c558d","source":"semantic_scholar","pdf_url_source":"semantic_scholar_open_access_pdf","license":"CCBY","status":"GOLD","reason":null},"reference_availability":{"status":"available","references_indexed":true,"full_text_available":false,"full_text_source":null,"count_basis":"semantic_scholar_metadata","extraction_status":"not_applicable","reason":null},"source":{"provider":"episteme2","base_corpus":"semantic_scholar_dump","freshness_mode":"unknown","basis":["semantic_scholar_metadata","postgres_metadata"],"limits":["paper metadata is based on indexed upstream scholarly datasets","claims and concepts are available only for extracted papers","absence of claims or concepts means no extracted graph data is available in this response"],"status":"available","degraded":false,"degraded_reasons":[],"diagnostics":{"status":"available","degraded":false,"degraded_reasons":[],"metadata_status":"available","graph_status":"available","abstract_status":"available"},"source_flags":1},"paper_id":630735,"paper_uid":"8b71b0dd-2bcb-4725-84d3-8c650c6d0a13","canonical_identity":{"paper_id":630735,"paper_uid":"8b71b0dd-2bcb-4725-84d3-8c650c6d0a13","identity_status":"available","lookup_basis":"semantic_scholar_external_id","compatibility_path":"corpus_id"},"url":"https://sah.borca.ai/papers/191193864"}