Abstract
Device-to-Device (D2D) communication is a key feature of IMT-Advanced systems, allowing direct communication between user equipments (UEs) by sharing the authorization band. However, resource sharing introduces interference. This paper analyzes a scenario where D2D UEs reuse uplink resources allocated to cellular users (CUs), creating aggregate interference at the base station (eNB). Using random geometry and information theory, the authors derive the quantitative relationship between the CU's successful communication radius, D2D user density, and cellular throughput, subject to minimum throughput requirements and interruption probability constraints.
Introduction
As wireless spectrum becomes increasingly crowded, D2D technology offers a way to improve data rates and reduce power consumption by bypassing the base station for local traffic. However, when D2D pairs operate in the same spectrum as a cellular user's uplink, they create an "interference soup" at the eNB. A critical challenge for network design is determining the "safe zone" or available communication radius for a cellular user. If a CU is too far from the eNB while surrounded by active D2D pairs, its signal will be drowned out by aggregate interference, leading to communication failure.
System Model and Methods
The study employs a theoretical model of a single cell with an eNB at the center. D2D transmitters are assumed to be randomly distributed according to a specific density ($\lambda$) and reuse the CU's uplink resources.
Key components of their approach include:
- Path Loss Modeling: They use a standard power-law path loss function ($1/d^\alpha$) to calculate signal decay.
- Random Geometry: Instead of looking at a single interferer, they use the density of D2D users ($\lambda$) to calculate the expectation of aggregate interference at the base station.
- Information Theory: They employ the Shannon formula to calculate throughput and use the Markov inequality to derive the probability of "communication interruption"—the chance that throughput falls below a required threshold ($C_0$).
Key Insights and Results
The core finding is the "Shrinking Radius" effect: there is a direct trade-off between D2D popularity and cellular range. As the density of D2D users ($\lambda$) increases, the "successful communication radius" of the cellular user must decrease to maintain the same quality of service.
Key results from the simulation include:
- Location Sensitivity: Interference is highly non-linear. A D2D transmitter very close to the base station is significantly more damaging to system throughput than multiple users at the cell edge.
- The Power Tug-of-War: To maintain a large communication radius in a D2D-heavy environment, cellular users must increase their transmission power ($P_{CU}$). However, this has diminishing returns and eventually hits a ceiling defined by the system's outage probability.
- Aggregate Interference is Key: The paper shows that the sum of many small D2D signals creates a noise floor that fundamentally limits the base station's "hearing" range.
- The Master Equation: The authors derived a closed-form inequality that allows network engineers to calculate the maximum allowable radius $R$ based on a given D2D density and a required interruption probability.
Limitations and Future Work
While the model provides a strong theoretical foundation, it has limitations:
- Single Cell/Single CU: The model assumes only one cellular user is active in the uplink at a time. Real-world scenarios involving multiple CUs across multiple cells would introduce more complex inter-cell interference.
- Static Environment: The paper assumes a random but static distribution of users and does not account for high mobility or fast-fading effects.
- Uplink Only: The study focuses exclusively on the uplink period. A holistic view would need to incorporate downlink interference, where the base station interferes with D2D receivers.
Summary updated