Name for a circular data structure delivery system? - data-structures

Let's say I have a data structure, which is essentially a circularly linked list. The usage of this circularly linked list is to continuously be walked, and at each node, delivery the data at that node to consumers. Therefor, the more frequently the same thing appears in the circularly linked list, the more frequently it will be delivered to a consumer.
Is there a name for this data structure?

Circular queue, circular buffer, cyclic buffer or ring buffer. I think circular buffer is the most common name I've heard.

Ringbuffer.

Related

Why can't a priority queue wrap around like an ordinary queue?

I know that in order to improve efficiency, Queues use the wrap around method, to avoid to move everything down all the time that we delete an element.
However, I do not understand why Priority Queues can't wrap around like ordinary Queues. In my point of view, Priority Queues have more similar behaviour to Stack than to a Queue, how is it possible?
The most common priority queue implementation is a binary heap, which would not benefit from wrapping around. You could create a priority queue that's implemented in a circular buffer, but performance would suffer.
It's important to remember than priority queue is an abstract data structure. It defines the operations, but not the implementation. You can implement priority queue as a binary heap, a sorted array, an unsorted array, a binary tree, a skip list, a linked list, etc. There are many different ways to implement a priority queue.
Binary heap, on the other hand, is a specific implementation of the priority queue abstract data type.
As for stack vs queue: in actuality, stacks and queues are just specializations of the priority queue. If you consider time as the priority, then what we call a queue (a FIFO data structure), is actually a priority queue in which the oldest item is the highest priority. A stack (a LIFO data structure) is a priority queue in which the newest item is the highest priority.

Queue with memory limited size

I need a queue with next properties and supported operations:
Push element in the beginning, automatically popping out from the end until size of queue become not greater than predefined limit.
Take N elements from beginning lazily and not traversing whole queue.
Limit by total size of all elements like 2Mb.
I know I can implement this by myself as a wrapper aroud Data.Sequence, or something else (according to mentioned implementations). Also found this blog post from Well Typed. I just wonder is this already implemented somewhere?
And if there is no standard implementation with desired behaviour, it would be nice to hear recommendations which standard data structure to use to implement such queue.
There is lrucache library which has almost anything I want except it limits it's size by number of elements in queue.

What is meant by re-buffering issue in a dequeue operation

I was going through a circular queue post, and it mentioned about re-buffering problem in other queue datastructures.
In a standard queue data structure re-buffering problem occurs for each dequeue operation. This problem can be solved by joining the front and rear ends of a queue to make the queue as a circular queue.
Circular queue is a linear data structure. It follows FIFO principle.
Can someone explain me what is re-buffering problem and how it happens during a dequeue operation ?
In a standard queue, implemented using array, when we delete any element only front is increment by 1, but that position is not used later. So when we perform many add and delete operations, memory wastage increases. But in Circular Queue , if we delete any element that position is used later, because it is circular.
This re-buffering problem occurs if queue is implemented using array. A circular queue implemented using array does not have re-buffering issue for dequeue operation.

Routing Table implementation : linked list vs array

I am trying to implement my own routing protocol over BTLE PHY and link layers to have a multi-hop link for BTLE radio. I am using Cortex-M0 processor for the same. My routing table structure is basically as follows:
|Neighbour Address| Info about Link quality | Possible Destination Addr|
The neighbour address will have address of immediate neighbour and possible destination addresses field will have addresses of destinations (within one hop) that can be reached from that particular neighbour (the routing only supports 2-hop communication). In short, the possible destinations will have entries of the elements which are in Neighbour address of the neighbour.
I am implementing this in C with CodeSorcery Toolchain Bare for ARM. So, for building a routing table, should a linked list be used or an array? Using array will be easier than implementing a linked list but then, the size of array will be predefined and limited. Plus, when initializing, it will eat up all the space dedicated to it. Will it be actually good to reserve space for routing table so that it will not cause memory problem later? Or should it be linked list which is more flexible in data allocation?
The best data structure is to do both. You can allocate address tables in 'blocks'. Each block will contain multiple routing table entries and the next pointer. After you fill up the block you can allocate the next block and fill it's address in the previous block next pointer.
Such structure gets both of both worlds, you get the speed of simple table scan and flexibility of the linked list.
How large your network is going to be? For very large networks, simple table will be a performance bottleneck, so you should consider using a hash table of prefixes so that you don't need to traverse the entire table to find a particular neighbor.

Are Priority Queues really Queues?

In Priority Queues, an element is inserted and deleted from the queue according to its priority, and because of which while writing the insertion and deletion code of elements for any priority queue; insertion and deletion are done according to the priority of the elements.
Suppose you have a queue with elements 1,5,6 and the priority of the elements is the value of the elements itself, and now one needs to insert an element of priority 3; then the elements is inserted at the second location in queue giving the new queue 1,3,5,6.
But a queue is defined as a data structure in which elements can be inserted at end and deleted at beginning but not in the middle, but in the above described case element is inserted at the second location (that is in the middle of queue). So if priority queues not obeying definition of queue so Are Priority Queues really Queues?
Kindly explain.
Priority queues are "queues" in one sense of the word, in that elements wait their turn. They are not a subtype of the Queue abstract data type.
A queue is characterized as an information structure in which components might be embedded at closure and erased at starting yet not in the center, however in the above portrayed case component is embedded at the second area (that is amidst queue).
Yes, a priority queue is still a queue in the sense that items are being served in the order in which they are located in the queue. However, in this case a priority is associated with each item and they are served accordingly.
A priority queue is a queue in the sense of the English word queue, not as a strict subtype of the other data structure named 'queue'. There is no inheritance going on there, they're just names that describe their purpose.

Resources