This study analyzes Tamil Nadu telephone data per 100 people from 2004 to 2021 and forecasts trends for the following five years using the ARIMA model. The dataset includes Tamil Nadu's 2004–2021 telephone connections per 100 people. The Augmented Dickey-Fuller (ADF) test checks data stationarity before analysis. Non-stationary data can cause erroneous forecasts in time series analysis. The ADF test will detect if data needs differencing for stationarity. To determine ARIMA model orders, the ACF and PAF will be investigated. These functions provide data point correlations at different lags, helping pick ARIMA parameters. The ARIMA (0,2,1) model fits telephone data best based on ACF and PACF graphs. To achieve stationarity, the data must be differenced twice, and the model will have one moving average component lag. The ARIMA (0,2,1) model has AIC and BIC values of 126.19 and 127.73, respectively. Lower numbers imply greater model fit. The ARIMA (0,2,1) model predicts 2022–2025 telephone statistics. This projection will reveal Tamil Nadu's five-year telephone connection trend per 100 people. Finally, the Box test will evaluate ARIMA model goodness-of-fit. The Box test validates model predictions by detecting considerable residual autocorrelation. This detailed research of Tamil Nadu's telephone data patterns will assist stakeholders make educated decisions about communications infrastructure and services for the future.