Class LoadDependentExamples
-
Constructor Summary
Constructors -
Method Summary
Modifier and TypeMethodDescriptionstatic voidFlow-equivalent-server aggregation of a two-class tandem (fes_aggregation).static voidFlow-equivalent-server aggregation of a single-class tandem (fes_single_class).static voidDemonstrates load-dependent behavior with class dependence (ld_class_dependence.ipynb).static voidThe same aggregation with two classes (ld_fes_multiclass).static voidNorton's theorem on a single-class tandem, solved by convolution (ld_fes_singleclass).static voidDemonstrates joint (non-product-form) load dependence (ld_joint_dependence.m).static voidDemonstrates load-dependent multi-server with FCFS discipline (ld_multiserver_fcfs.ipynb).static voidDemonstrates load-dependent multi-server with processor sharing (ld_multiserver_ps.ipynb).static voidDemonstrates load-dependent PS with two classes (ld_multiserver_ps_twoclasses.ipynb).static voidMain method to run all load-dependent examples.
-
Constructor Details
-
LoadDependentExamples
public LoadDependentExamples()
-
-
Method Details
-
ld_class_dependence
Demonstrates load-dependent behavior with class dependence (ld_class_dependence.ipynb). This example shows how service rates can depend on the mix of customer classes present in the queue. This models situations where different types of customers create different levels of resource contention. Features: - Service rates dependent on class composition - Multi-class load-dependent queues - Analysis of class interaction effects - Performance isolation between classes- Throws:
Exception- if the solver encounters an error
-
ld_joint_dependence
Demonstrates joint (non-product-form) load dependence (ld_joint_dependence.m).The station rate reads the class-1 marginal only and is shared by every class, so it is NOT the product-form beta_{i,r}(n_{i,r}) of
ld_class_dependence(); the exact CTMC is the reference against which the QD-AMVA approximation is read.- Throws:
Exception- if the solver encounters an error
-
ld_multiserver_fcfs
Demonstrates load-dependent multi-server with FCFS discipline (ld_multiserver_fcfs.ipynb). This example models a multi-server queue where the effective service rate depends on the number of customers in the system. Under FCFS, this creates complex interactions between waiting times and system occupancy. Features: - Multi-server queue with load-dependent rates - FCFS scheduling discipline - Modeling of server activation/deactivation - Analysis of economies/diseconomies of scale- Throws:
Exception- if the solver encounters an error
-
ld_multiserver_ps
Demonstrates load-dependent multi-server with processor sharing (ld_multiserver_ps.ipynb). This example shows load-dependent behavior in a processor sharing environment. PS discipline distributes service capacity equally among all customers, and the load dependence models how this capacity changes with system load. Features: - Multi-server queue with processor sharing - Load-dependent service capacity - Fair sharing of variable capacity - Comparison with FCFS load dependence- Throws:
Exception- if the solver encounters an error
-
ld_multiserver_ps_twoclasses
Demonstrates load-dependent PS with two classes (ld_multiserver_ps_twoclasses.ipynb). This example extends the PS load-dependent case to multiple customer classes. Each class can have different impacts on the load-dependent behavior, modeling heterogeneous workloads with different resource requirements. Features: - Two-class processor sharing system - Class-specific load dependence effects - Analysis of fairness across classes - Performance differentiation strategies- Throws:
Exception- if the solver encounters an error
-
fes_aggregation
Flow-equivalent-server aggregation of a two-class tandem (fes_aggregation).Queue1 and Queue2 are replaced by one limited-class-dependent station whose per-class rates are the isolated subnetwork's throughputs. For a closed product-form network Norton's theorem makes this EXACT, so the aggregated throughput reproduces the original's rather than approximating it, which is what the comparison below prints.
- Throws:
Exception- if the solver encounters an error
-
fes_single_class
Flow-equivalent-server aggregation of a single-class tandem (fes_single_class).- Throws:
Exception- if the solver encounters an error
-
ld_fes_singleclass
Norton's theorem on a single-class tandem, solved by convolution (ld_fes_singleclass).All THREE queues are aggregated, leaving the think time beside one FES, and the aggregate is solved with exact NC: the FES rates are Sauer's chain-dependent service rates, which the convolution consumes directly.
- Throws:
Exception- if the solver encounters an error
-
ld_fes_multiclass
The same aggregation with two classes (ld_fes_multiclass).- Throws:
Exception- if the solver encounters an error
-
main
Main method to run all load-dependent examples.- Parameters:
args- command line arguments (not used)
-