⚡ electrical engineering

An Enhanced Time Space Priority Scheme to Manage QoS for Multimedia Flows transmitted to an end user in HSDPA Network

This paper proposes and mathematically models an Enhanced Basic-Time Space Priority (EB-TSP) scheme to overcome the limitations of existing combined time and space priority methods, thereby improving Quality of Service for real-time multimedia flows and optimizing network resource utilization in HSDPA networks.

Original authors: Mohamed Hanini, Abdelali El Bouchti, Abdelkrim Haqiq, Amine Berqia

Published 2026-06-04

📖 4 min read☕ Coffee break read

CC BY 3.0

Original authors: Mohamed Hanini, Abdelali El Bouchti, Abdelkrim Haqiq, Amine Berqia

Original paper licensed under CC BY 3.0 (http://creativecommons.org/licenses/by/3.0/). ✨ This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

Imagine a busy highway leading into a city (the HSDPA network). On this highway, two types of vehicles are trying to get to their destination: Emergency Ambulances (Real-Time packets like voice and video calls) and Delivery Trucks (Non-Real-Time packets like emails and web browsing).

The city has a limited number of parking spots (the buffer) before the main gate. The goal is to make sure the Ambulances get through quickly without getting stuck, while still letting the Trucks get in eventually without losing their cargo.

The Problem: The Old "Strict" Rule

The paper discusses an old traffic rule called Basic-TSP. Here is how it worked:

Time Priority: Ambulances always get to drive in front of Trucks.
Space Priority (The Flaw): However, there was a strict rule: "Only the first 15 Ambulances are allowed to park in the special front section."

The Catch: If the front section was full (15 Ambulances), and a 16th Ambulance arrived, the gate would say, "No more Ambulances allowed!" Even if there were empty spots in the back section reserved for Trucks, the 16th Ambulance would be turned away and dropped. Meanwhile, the Trucks were stuck behind the 15 Ambulances, waiting their turn.

This was inefficient. It wasted empty parking spots and caused valuable Ambulance calls to be dropped unnecessarily.

The Solution: The "Enhanced" Rule (EB-TSP)

The authors propose a new, smarter traffic rule called Enhanced Basic-TSP (EB-TSP).

Here is how the new system works:

Ambulances still go first: They always drive in front of the Trucks.
The "Smart Swap": If the front section is full (15 Ambulances) and a new Ambulance arrives, the system checks the back section. If there is a Truck parked there, the system says, "Okay, we have an empty spot in the back. We will move the Truck out (drop it) to make room for this new Ambulance."

The Analogy: Think of it like a VIP line at a club.

Old Rule: The VIP line has a strict limit of 15 people. If a 16th VIP arrives, they are turned away, even if the General Admission line is empty.
New Rule: The VIP line still has priority. If the VIP line is full, the bouncer looks at the General Admission line. If there is someone there, the bouncer says, "Sorry, General Admission, you have to leave so this VIP can get in." This ensures VIPs rarely get turned away, while still managing the total crowd size.

What the Math and Numbers Show

The authors built a mathematical model (like a computer simulation of traffic patterns) to test this new rule. They compared the "Old Rule" (Basic-TSP) with the "New Rule" (EB-TSP).

The Results:

Fewer Dropped Calls: The new rule significantly reduced the number of Ambulances (RT packets) that were turned away. This means fewer dropped voice and video calls for the user.
Same Performance for Trucks: The number of Trucks (NRT packets) that were dropped or delayed stayed roughly the same as before. The new rule didn't make things worse for the general data traffic; it just made the system smarter about using empty space.
Better Resource Use: The system stopped wasting empty parking spots.

The Bottom Line

The paper concludes that by simply changing the rule on how to handle a full queue—allowing the system to swap a lower-priority packet for a higher-priority one when space is available—you can significantly improve the quality of real-time services (like video calls) without hurting the performance of data services. It's a smarter way to manage the "parking lot" of the mobile network.

Technical Summary: An Enhanced Time Space Priority Scheme to Manage QoS for Multimedia Flows in HSDPA Networks

Problem Statement
In High-Speed Downlink Packet Access (HSDPA) networks, end users often receive multiple concurrent data flows with divergent Quality of Service (QoS) requirements. Real-Time (RT) packets (e.g., voice, video) require bounded delay, while Non-Real-Time (NRT) packets (e.g., web browsing, email) are sensitive to packet loss. When these flows share limited network resources, specifically buffer space and bandwidth, effective queue management and scheduling are essential.

The paper identifies a specific limitation in the existing "Basic Time Space Priority" (B-TSP) scheme. In B-TSP, RT packets are given transmission priority (time priority) but are restricted by a threshold $R$ regarding the maximum number of RT packets allowed in the buffer (space priority). The authors argue that this rigid threshold creates a significant drawback: if the buffer is full and the number of RT packets is less than $R$ , an arriving RT packet is still rejected if the buffer is occupied by NRT packets. This results in unnecessary RT packet drops even when buffer space is technically available (reserved for NRT), leading to suboptimal QoS for RT flows and inefficient buffer utilization.

Methodology
To address the limitations of B-TSP, the authors propose an "Enhanced Basic-TSP" (EB-TSP) scheme. The core mechanism modifies the admission control logic within the buffer:

Admission Logic: When an RT packet arrives, if the buffer is full but the current number of RT packets is below the threshold $R$ , the system rejects an NRT packet to admit the RT packet. Conversely, if an NRT packet arrives and the buffer is full, it is rejected only if the number of RT packets is below $R$ ; otherwise, an RT packet is rejected to admit the NRT packet.
Service Logic: RT packets always remain at the front of the queue for transmission (time priority), ensuring they are served before NRT packets.

To evaluate the scheme, the authors develop a mathematical model using a continuous-time Markov chain (CTMC).

Assumptions: RT and NRT packet arrivals follow independent Poisson processes ( $\lambda_{RT}$ , $\lambda_{NRT}$ ), and service times follow independent exponential distributions ( $\mu_{RT}$ , $\mu_{NRT}$ ).
State Space: The system state is defined by the number of RT and NRT packets in the buffer.
Analysis: The authors derive equilibrium probabilities by solving a system of balance equations. Using these probabilities, they analytically calculate key QoS metrics: loss probability for both RT and NRT packets, the average number of packets in the buffer, and mean delay (derived via Little's Formula).

Key Contributions

Algorithmic Enhancement: The proposal of the EB-TSP algorithm, which dynamically manages buffer space to prevent the rejection of RT packets when NRT packets occupy space below the RT threshold $R$ .
Mathematical Modeling: The formulation of an ergodic Markov chain model to analytically characterize the transition of the system under the EB-TSP scheme, providing closed-form expressions for QoS measures.
Comparative Analysis: A rigorous comparison between the proposed EB-TSP and the standard B-TSP scheme, demonstrating that the enhancement specifically targets RT packet loss without degrading other performance metrics.

Results
Numerical simulations were conducted using Maple software with a total queue length of 60, an RT threshold of 15, and specific arrival/service rates. The results indicate:

RT Packet Loss: The EB-TSP scheme significantly reduces the loss probability of RT packets compared to B-TSP, particularly as the arrival rate of RT packets increases.
RT Delay: The average delay for RT packets remains comparable between the two schemes.
NRT Performance: The loss probability and average delay for NRT packets in the EB-TSP scheme remain at the same level as in the B-TSP scheme.

Significance
The paper claims that the EB-TSP scheme offers a more efficient management of network resources in HSDPA environments. By overcoming the specific drawback of the B-TSP scheme where RT packets are dropped despite available buffer capacity, the proposed method improves the perceived quality for end users, specifically for audio and video calls. The authors conclude that the scheme achieves a better balance, enhancing QoS for RT flows while maintaining the performance levels of NRT flows and ensuring effective buffer space utilization.

The Problem: The Old "Strict" Rule

The Solution: The "Enhanced" Rule (EB-TSP)

What the Math and Numbers Show

The Bottom Line

More like this