Multi-Server Queues: Priorities and Jackson's Theorem
This paper examines multi-server queuing theory, covering how multiple servers share a queue and dispatch items according to various service disciplines. It introduces the concept of multiple priority classes, explaining how jobs are labeled and served according to priority schemes—including preemptive service and Service Level Agreements (SLAs)—and why priority analysis is more complex in multi-server settings. The paper then presents Jackson's Theorem, which provides a product-form expression for the joint state distribution of a network of queues, and outlines the key implications of the theorem: flow balance, queue independence, and Poisson flow behavior within queue networks.
- Introduction to Multi-Server Queues: Definition, mechanics, and history of multi-server queues
- Multiple-Server Queues and Priority Classes: Priority schemes, SLAs, and complexity in multi-server settings
- Jackson's Theorem and Its Assumptions: Product-form theorem and required distributional assumptions
- Implications of Jackson's Theorem: Three key consequences for queue independence and Poisson flows
- References: Primary sources in queuing theory and priority systems
✍️ How to write this paper — guide, tools & examples ▾
What makes this paper effective
- Clearly defines foundational concepts (multi-server queues, priority classes, SLAs) before introducing the more abstract theorem, making the material accessible.
- Uses direct quotations from primary sources to support technical claims, grounding the argument in established queuing theory literature.
- Enumerates the three implications of Jackson's Theorem in a structured list, aiding comprehension of a mathematically complex result.
Key academic technique demonstrated
The paper demonstrates disciplined use of synthesis: it draws on multiple technical sources (Harchol-Balter et al., Bose, and an NSF workshop) and weaves their findings into a coherent progression from basic definitions through priority systems to the formal theorem. This technique — building from concrete application to abstract formulation — is effective in applied mathematics and operations research writing.
Structure breakdown
The paper opens with a definition of multi-server queues and their historical context, then moves to priority classes and the challenges they introduce in multi-server settings, and concludes with Jackson's Theorem, its assumptions, and its three stated implications. The argument follows a logical build-up: concept → complication → theoretical resolution.
Introduction to Multi-Server Queues
A multi-server queue is a queue shared by multiple servers. When an item arrives and at least one server is available, the item is dispatched immediately to that server. The assumption is made that all servers are identical; therefore, it makes no difference which server is chosen for the item.
When all servers are busy, the queue begins to form, and upon one server becoming free, the next item is dispatched from the queue according to the dispatch discipline being used. Other than the parameter of utilization, all other parameters carry over to the multi-server case with the same interpretation.
The study of queues with multiple servers has been ongoing for approximately fifty years and dates back to the work of Kiefer and Wolfowitz. Their work resulted in enhancements to call centers, checkout lines, and high-performance computing systems, since multiple servers enable increased performance and are "cost-effective and easily scalable" (NSF & GSIA, 2004).
Multiple-Server Queues and Priority Classes
Harchol-Balter, Osogami, Scheller-Wolf, and Wierman (2003) note that a great deal of queuing theory is devoted to the analysis of priority queues, in which jobs or customers are "labeled and served in accordance with a priority scheme: high-priority jobs preempt medium-priority jobs, which in turn preempt low-priority jobs in the queue." In fact, many real-world applications involve priority queuing.
Sometimes a job's priority is determined according to the job owner's status through what is known as a Service Level Agreement (SLA), such as when customers choose to pay a higher amount in order to receive "high-priority access to some high-demand resource" (Harchol-Balter et al., 2003). At other times, a job's priority is assigned artificially in order to "maximize a company's profit or increase system utilization" (Harchol-Balter et al., 2003).
Analyzing the mean response time and higher moments of response time for various classes of jobs is important, and this analysis is more difficult in the context of a multi-server system. Adding to the complexity is the fact that multi-server systems are "prevalent in many applications where prioritization is used" (Harchol-Balter et al., 2003). The difficulty arises because jobs of different priorities may be in service at the same time, complicating the mathematical treatment of the system.
Create your account
Always verify citation format against your institution’s current style guide requirements.