Ensuring food security in precision agriculture demands early prediction of corn yield in the United States of America at international, regional, and local levels. Accurate yield estimation can play a crucial role in averting famine by offering insights into food availability during the growing season. To address this, we proposed a Concatenate-based 2D-CNN-BILSTM model that integrates Sentinel-1, Sentinel-2, and SoilGRIDs (global gridded soil information) data for corn yield estimation in Iowa State from 2018 to 2021. Sentinel-2 images provided essential bands and indices, such as Blue, Green, Red, and Red Edge 1/2/3/4, NIR, and SWIR 1/2 Bands, along with NDVI, LSWI, DVI, RVI, DWRVI, SAVI, VARIGREEN, and GNDVI. Additionally, VV, VH, difference VV, and VH, and RVI were extracted from Sentinel-1 SAR images. Soil data encompassing silt, clay, sand, cec, and pH were collected at various depths ranging from 0 cm to 200 cm. To extract high-level features for each month, a dedicated 2D-CNN was designed. The 2D-CNN concatenated high-level features from the previous month with low-level features of the subsequent month, serving as input features for the 2D-CNN. To incorporate single-time soil data features, another 2D-CNN was created. Finally, the high-level features from soil, Sentinel-1, and Sentinel-2 data were concatenated and fed into a BILSTM layer to predict corn yield. The performance of the proposed Concatenate-based 2D-CNN-BILSTM model was compared against random forest (RF), Concatenate-based 2D-CNN, and 2D-CNN models using some metrics like RMSE, MAE, MAPE, and the Index of Agreement. The model was trained on data from 2018, 2019, and 2020, and its accuracy was tested using data from 2021. Results revealed the Concatenate-based 2D-CNN-BILSTM model's impressive Index of Agreement of 84.67% and low RMSE of 0.698 t/ha. Furthermore, validation demonstrated high prediction accuracy for the Concatenate-based 2D-CNN model, with an RMSE of 0.799 t/ha and an Index of Agreement of 72.71%. The 2D-CNN model also performed well, yielding an RMSE of 0.834 t/ha and an Index of Agreement of 69.90%. However, the RF model exhibited lower accuracy with an RMSE of 1.073 t/ha and an Index of Agreement of 69.60%. These findings highlight the potential of utilizing advanced deep-learning techniques in conjunction with remote sensing and soil data to enhance crop yield predictions. The integration of Sentinel 1-2 and SoilGRIDs data with the proposed 2D-CNN-BiLSTM model demonstrated significant improvements in accuracy for corn yield prediction. The combination of soil data and features extracted from Sentinel 1-2 resulted in a decrease in RMSE by 16 kg and an increase in the Index of Agreement by 2.59%. This study demonstrates the power of leveraging advanced Machin Learning (ML) methods for achieving accurate and reliable predictions to support sustainable agricultural practices and food security initiatives.