The Internet of Things (IoT) is a network of interconnected, Internet-connected items that can gather and transmit data without the need for human interaction through a wireless network. Cryptography is a way of using codes to safeguard information and communications so that only those who are supposed to read and process it may do so. Existing system shows privacy and replication issues caused during cipher text attribute set based encryption. The technical issues in achieving secured authentication is removed by using multi-channel attack detection based on lightweight message authentication code (LMAC) access control using IoT design in which collision attack can be removed by using secure hash algorithm based Merkle tree and to control sinkhole attacks agent navigation algorithm is proposed and long short-term memory (LSTM) recurrent neural network (RNN) is proposed to provide the conformity of the user login password. Moreover, to identify attackers on the transmitter side by ensuring authentication, multichannel detection method is proposed. Finally, the security and privacy is improved by using LMAC and user revocation. Thus the multi-channel attack detection based on LMAC using IoT design can be outperforming the other existing model with higher throughput and low execution time.