Supporting data and code for dual-frequency superposition-based vibration communication and carrier screening in coupled tubing strings
Description
This dataset supports the study “Dual-frequency superposition-based vibration communication and an associated carrier-screening method for coupled tubing strings.” The study tests the hypothesis that periodic couplings, finite tubing-string length, and the finite spectral bandwidth of short symbols jointly cause strong frequency selectivity, so carriers selected only from short-string response peaks may become weak when the number of tubing–coupling periods changes. The dataset contains: (1) a cleaned Code Composer Studio 12.4 project for the Texas Instruments TMS320F28335, with Example_2833xAdcSoc.c as the main source file; (2) MATLAB_Online_Decoder.m for SCI reception, frame reconstruction, online decoding, display, and storage of 15 predefined decimal-value groups; (3) Unified_Periodic_Response_Data.csv, containing processed finite-period response data for N = 1, 3, 6, and 9; and (4) Full_Band_Carrier_Robustness_Results.csv, containing spectrally weighted responses and carrier-screening metrics. The finite-period data were obtained from frequency-domain models of coupled tubing strings and arranged on a common frequency grid. The carrier-screening results were calculated from the effective response covered by the main lobe of a finite-duration symbol. For each carrier, G_worst is the minimum effective response across N = 1, 3, 6, and 9, R is the maximum-minus-minimum response range, and S_lambda = G_worst - lambda R is the robustness score. Higher G_worst and S_lambda indicate better worst-case response and overall robustness, while a smaller R indicates lower sensitivity to period count. The data reveal recurring broadband low-response regions that generally become more pronounced as period count increases. Under the evaluated 26 ms symbol condition, the results identify weak-carrier regions and provide candidate replacement frequencies ranked by worst-case response and cross-period stability. These rankings are screening results only. They do not directly represent communication capacity, delay, intersymbol interference, dual-frequency balance, or decoding accuracy. The DSP code generates dual-frequency waveforms, outputs them through the DAC5620C, acquires vibration-response samples through the ADC, and transfers 64-sample blocks through SCI-A at 468750 baud using the synchronization header A5 5A 3C C3. MATLAB performs online reception and decoding. Reuse with different tubing lengths, coupling counts, sensor mounting conditions, amplifier settings, or acquisition chains requires recalibration of carrier frequencies, equalization coefficients, and decision thresholds. The complete original raw ADC waveform archive is no longer retained. The available processed response data, carrier-screening results, DSP source code, MATLAB program, and README are provided to support interpretation and partial reproduction of the reported workflow.
Files
Steps to reproduce
1. Download all files and extract CCS_Source_Code.zip. 2. Open Code Composer Studio 12.4 and import the extracted CCS project by selecting File > Import > Code Composer Studio > CCS Projects. 3. Confirm that the target device is the Texas Instruments TMS320F28335 and that Example_2833xAdcSoc.c is included as the main DSP source file. 4. Check the target configuration, compiler settings, include paths, and linker command files. Build the project and load it onto the TMS320F28335 after connecting the experimental transmission and acquisition hardware. 5. The DSP program generates the synchronous dual-frequency waveform, outputs it through the DAC5620C, acquires the vibration response through the ADC, and transfers 64-sample data blocks through SCI-A. The SCI settings are 468750 baud and 8N1. Each data block is preceded by the synchronization header A5 5A 3C C3. 6. Open MATLAB_Online_Decoder.m in MATLAB. Set the serial-port identifier to the port connected to the DSP and confirm the baud rate and synchronization-header settings. 7. Run MATLAB_Online_Decoder.m, and then start the DSP program. MATLAB receives the SCI data, reconstructs the ADC blocks, searches candidate frame phases, performs frequency-domain character recognition, checks the five-character decimal format, and displays and stores the decoded results for the 15 predefined decimal-value groups. 8. Open Unified_Periodic_Response_Data.csv to examine the finite-period response data for N = 1, 3, 6, and 9. 9. Open Full_Band_Carrier_Robustness_Results.csv to examine the spectrally weighted effective responses, worst-case response, across-period response range, and carrier-robustness scores. 10. Carrier frequencies, empirical equalization coefficients, decision thresholds, serial-port identifiers, and local storage paths should be recalibrated or modified when the tubing configuration, sensor mounting, amplifier setting, acquisition chain, or computer environment changes. The complete original raw ADC waveform archive is no longer retained. The available processed response data, carrier-screening results, DSP source code, and MATLAB online-decoding program are provided.
Institutions
- Liaoning Shihua UniversityLiaoning, Fushun