This work introduces novel random projection-based efficient time complexity detectors for an uplink massive multiple-input multiple-output (MIMO) communication networks. The proposed random projection-based detectors reduce the original dimension of the received symbols while preserving the pairwise Euclidean distance between the original and the compressed dimensions with high probability. Consequently, obtaining a faster detection algorithm with a comparable detection performance. Building on several variants of random projection such as Rademacher, Very Sparse Random Projection (VSRP), and Fast Johnson Lindenstrauss Transform detectors (FJLT), the corresponding detectors, ŝ RP-ZF , ŝ VSRP and ŝ FJLT are presented. A closed-form expression of the approximate symbol error probability (SEP) is obtained to characterize their performance. The time complexity of the proposed detectors is shown to significantly improve from the benchmark detectors, namely the maximum likelihood (ML), zero-forcing (ZF), and minimum mean-squared error (MMSE) detectors, along with the class of reduced complexity Neumann-series-based matrix inverse approximation (NS-MIA) detectors. The simulation results validate the tradeoff between the detection performance and the time complexity of the random projection-based detectors.