Abstract:Task offloading plays a pivotal role in mobile edge computing, enabling terminal devices to enhance task execution efficiency and conserve energy. However, servers are reluctant to offer services without compensation. Currently, pricing mechanisms are commonly employed to incentivize servers to serve terminal devices, with servers earning revenue through payments from these devices. Given the rapid surge in terminal devices, determining the optimal number of servers placement for service providers (SPs) to maximize revenue is crucial. In this paper, we propose a server placement scheme based on an all-pay auction framework. Experimental simulations reveal that an optimal server-user ratio of approximately 25% maximizes SP profits.
What problem does this paper attempt to address?
The problem that this paper attempts to solve is how to determine the optimal number of server deployments in the Mobile Edge Computing (MEC) environment to maximize the revenue of Service Providers (SPs). Specifically, Task Offloading plays an important role in MEC, which can improve the task execution efficiency of end - devices and save energy. However, servers are unwilling to provide services for free, so a pricing mechanism is required to encourage servers to provide services for end - devices.
With the rapid growth of the number of end - devices, determining the optimal server - user ratio is crucial for SPs. If the number of servers is too small, it will lead to unmet demand and thus loss of revenue; if the number of servers is too large, it will lead to fixed costs for idle servers, thereby reducing the SPs' profit. Therefore, finding a balance point to maximize the total revenue of SPs is the core issue of this research.
To this end, the author proposes a server deployment scheme based on the All - pay Auction framework. Through experimental simulation, the author finds that when the server - user ratio is approximately 25%, the SPs' profit reaches the maximum.
### Main Contributions
1. **Propose a server deployment scheme based on All - pay Auction**: Use the All - pay Auction mechanism to encourage Edge Servers (ESs) to provide services for end - users.
2. **Determine the optimal server - user ratio**: Through simulation and theoretical analysis, it is determined that a server - user ratio of approximately 25% can maximize the SPs' revenue.
3. **Consider practical application scenarios**: Combine game theory and economic principles to ensure the feasibility and effectiveness of the proposed scheme in practical applications.
### Mathematical Models and Formulas
- **User Valuation Function**:
\[
F(v)=\frac{v}{A}
\]
where \(A = k\cdot\log_2(F_t)\), \(k>0\) is the valuation coefficient, and \(F_t\) is the user's computing power.
- **Equilibrium Bid**:
\[
b_e=\begin{cases}
\int_0^v t dF^{n - 1}(t),&n\geq2\\
e,&n = 1
\end{cases}
\]
where \(e\in[B+\log_2 q,u)\), \(u = v - e\), \(B\) is the fixed cost, and \(q\) is the amount of offloaded data.
- **Server Revenue**:
\[
U_j^{\text{server}}=\sum_{i = 0}^{n_i}b_i^e - B-\log_2(q_w)
\]
where \(q_w\) is the amount of data offloaded by the winner, and \(n_i\) is the total number of end - users in the same set.
- **Total Revenue**:
\[
W=\sum_{j = 0}^M U_j^{\text{server}}
\]
Through the derivation of these models and formulas, the author shows how to maximize the SPs' revenue by optimizing the server - user ratio and provides simulation results to support this conclusion.