Researchers from a collaborative international team have published a study in Discover Computing detailing the performance of an eight-channel wavelength-division multiplexing (WDM) optical communication system designed for Internet of Things (IoT) traffic. The study, led by Shippu Sachdeva and colleagues, investigates how varying traffic loads affect transmission quality over a 50-kilometer single-mode fiber link. The primary finding indicates that chromatic dispersion and inter-symbol interference are the dominant factors limiting data rates in such systems, with significant performance degradation observed at higher speeds despite the use of parametric optimization techniques.
The research addresses the exponential growth in data traffic driven by IoT-based applications, which require reliable, high-speed, and low-latency transmission. Traditional communication systems often struggle to handle the periodic, bursty, and heterogeneous nature of IoT traffic efficiently. To accommodate this large-scale transmission, the proposed work integrates WDM technology, which allows multiple signals to be transmitted simultaneously over a single optical fiber by using different wavelengths. This approach provides a more efficient solution for handling the increasing bandwidth demands of next-generation networks.
The experimental setup involved an eight-channel WDM system operating in the C-band, simulated using OptiSystem software. The system was tested under three varying traffic loads: 5 Gbps, 10 Gbps, and 15 Gbps. Each channel was generated using a pseudo-random binary sequence (PRBS) generator, which produces a binary data sequence corresponding to IoT traffic, followed by non-return-to-zero (NRZ) pulse shaping and continuous-wave lasers. The signals were multiplexed with a uniform spacing of 100 GHz and propagated through a 50-kilometer single-mode fiber link. An Erbium-Doped Fiber Amplifier (EDFA) with a gain of 20 dB was employed to compensate for signal attenuation during transmission.
To mitigate the effects of dispersion, a Dispersion Compensating Fiber (DCF) was incorporated into the system. The DCF provides negative dispersion to counteract the accumulated dispersion in the single-mode fiber. The researchers varied the length of the DCF to identify the optimal operating condition for different data rates. At the receiver side, the optical signal was processed through an optical Gaussian filter, a demultiplexer, a PIN photodetector, and a low-pass filter before being analyzed using a bit error rate (BER) analyzer. Key performance metrics included BER, Q-factor, optical spectra, and eye diagrams.
Baseline performance results showed that the system operated reliably at 5 Gbps, exhibiting the lowest BER and the highest Q-factor among the tested loads. At 10 Gbps, performance remained acceptable, although a slight reduction in the Q-factor was observed. However, at 15 Gbps, the BER increased significantly, and the Q-factor decreased, indicating a substantial impact on signal quality. The eye diagrams at 15 Gbps showed reduced opening and appeared nearly closed, proving severe noise and inter-symbol interference. This degradation is attributed to pulse broadening caused by chromatic dispersion over the 50-kilometer fiber link.
The study also examined the impact of laser launch power on system performance. The results indicated that negligible changes in BER and Q-factor were observed with variations in laser power. The Q-factor remained constant, suggesting that the system performance was not limited by insufficient power. Furthermore, increasing the optical power did not improve performance due to the onset of non-linear effects. This finding suggests that the system is not power-limited, and simply increasing launch power is not a viable strategy for enhancing transmission quality at higher data rates.
Similarly, the impact of EDFA gain was analyzed, and variations in gain did not significantly affect the BER or Q-factor. Although the EDFA compensates for fiber attenuation, increasing the gain also introduces amplified spontaneous emission (ASE) noise, which offsets any improvement in signal strength. The results indicate that the system is not amplification-limited, and increasing gain cannot overcome the degradation observed at higher data rates. This highlights that the primary constraints are related to dispersion and noise rather than signal power or amplification levels.
The effect of fiber length on performance was also investigated. The results showed a direct influence of transmission distance on operating data rates. At 5 Gbps, a local performance optimum was observed at 40 kilometers, with a Q-factor of approximately 17.1 and a BER decreasing to approximately 10^-6.5. As the link length increased to 50 and 60 kilometers, performance degraded due to accumulated dispersion and fiber loss. At 15 Gbps, a substantial reduction in the Q-factor was observed between 20 and 40 kilometers, followed by consistently poor performance at longer distances. The bit period decreases from 200 picoseconds at 5 Gbps to around 66.7 picoseconds at 15 Gbps, leaving less margin for waveform distortion and increasing sensitivity to inter-symbol interference.
Channel spacing was another critical parameter evaluated in the study. The best simulation performance was observed at 100 GHz channel spacing, where the Q-factor reached 13.5, 9.4, and 3.3 for 5, 10, and 15 Gbps, respectively. At 150 GHz, the Q-factor decreased to approximately 2–3 for all three data rates, indicating significant degradation. However, at 200 GHz, performance slightly improved, with Q-factors of approximately 9, 6.7, and 3.2 for 5, 10, and 15 Gbps. This non-monotonic behavior suggests that the relationship between channel spacing and performance is complex, depending on transmitter channel alignment, multiplexer and demultiplexer characteristics, and receiver response.
The study found that dispersion compensation using DCF significantly improved system performance, particularly at higher data rates. The optimal DCF length varied depending on the data rate. For 5 Gbps, the highest Q-factor of 17.8 was achieved at a DCF length of 8 kilometers. At 10 Gbps, the optimal length was also 8 kilometers, yielding a Q-factor of approximately 15.4. For 15 Gbps, the Q-factor increased from 3.1 at 2 kilometers to 13.1 at 10 kilometers, indicating that a longer DCF length was required to compensate for the greater dispersion effects at higher speeds. These results demonstrate that the optimal DCF length is data-rate dependent and must be adjusted to achieve the best performance.
The authors conclude that dispersion is the major factor limiting the data rate in the proposed WDM optical communication system. While parametric optimization of launch power and amplifier gain had negligible effects, adjusting the DCF length and channel spacing provided significant improvements in BER and Q-factor. The study provides valuable insights for the design of next-generation optical communication systems capable of handling high-data-rate IoT traffic. The findings suggest that dispersion management and spectral optimization are critical for enabling reliable transmission in WDM networks.
Limitations of the study include its focus on link-level simulation rather than a full network implementation. The traffic model uses pseudo-random binary sequences to represent IoT traffic, which does not fully capture the statistical behavior of real-world IoT networks, such as packet headers and burst arrivals. The study also does not account for advanced modulation formats or cross-layer optimization techniques that could further enhance performance. Future work could explore the use of reinforcement learning algorithms and neural networks to adaptively adjust system parameters, as well as the integration of WDM optical systems with 5G/6G networks for greater flexibility and robustness.
The mathematical framework underpinning the simulation defines the relationship between physical impairments and measurable performance metrics. Chromatic dispersion is modeled as the accumulation of pulse broadening, where the total accumulated dispersion is the product of the dispersion coefficient and the fiber length. In the proposed configuration, which combines single-mode fiber with dispersion-compensating fiber, the theoretical length of the compensating fiber is calculated to achieve first-order dispersion cancellation, where the total accumulated dispersion approaches zero. However, the study notes that the optimal dispersion-compensating fiber length for minimizing bit error rate and maximizing the Q-factor may differ from this theoretical value due to additional factors such as fiber attenuation, residual dispersion, amplifier noise, and receiver filtering. This distinction highlights that practical optimization requires evaluating the complete system response rather than relying solely on idealized cancellation conditions.
The modeling of Internet of Things traffic in this study relies on pseudo-random binary sequences to generate controlled digital data patterns. This approach allows the researchers to analyze the physical-layer link response to specific bit rates and pulse shapes without reproducing the full statistical behavior of a real-world IoT network. The study explicitly limits its scope to the link response, acknowledging that real IoT traffic includes packet headers, overhead packets, and burst arrivals that are not fully represented in the simulation. The traffic is mapped to wavelength-division multiplexing channels based on allocated priority and bandwidth, with higher-priority traffic assigned to lower-interference channels. This modeling choice ensures that the observed performance degradation is attributable to physical-layer impairments such as dispersion and inter-symbol interference, rather than network-layer congestion or protocol overhead.
The non-monotonic behavior observed in channel spacing and dispersion compensation underscores the complexity of optimizing multi-parameter optical systems. While increasing channel spacing generally reduces frequency overlap and inter-channel crosstalk, the study found that performance did not improve linearly with wider spacing. Similarly, the optimal length of the dispersion-compensating fiber varied significantly with data rate, requiring longer compensation for higher speeds. These findings suggest that parametric optimization must be performed holistically, considering the interaction between transmitter alignment, multiplexer and demultiplexer characteristics, and receiver response. The results indicate that isolated adjustments to single parameters, such as launch power or amplifier gain, are insufficient to overcome the fundamental limitations imposed by dispersion at higher data rates.
Subject of Research: Performance evaluation of IoT traffic in WDM optical communication systems
Article Title: Performance assessment of IoT traffic transmission in WDM optical communication systems with parametric optimization
Article References: Sachdeva, S., Chawla, S., Raghuwanshi, K. D., Kaur, J., Yajid, M. S. A., Chohan, M. K., & Shukla, M. K. (2026). Performance assessment of IoT traffic transmission in WDM optical communication systems with parametric optimization. Discover Computing, 29(1), Article 866. https://doi.org/10.1007/s10791-026-10836-0
Image Credits: AI Generated
DOI: 10.1007/s10791-026-10836-0
Keywords: IoT, WDM, Optical Communication, Dispersion Compensation, Bit Error Rate, Q-factor, Single-Mode Fiber, Parametric Optimization, Data Transmission, Network Performance, Performance, assessment
News Source: Denise Maddox. (November 9, 2026). Study Finds Dispersion Limits IoT Data in WDM Optical Networks. Scienmag.
