2024 5th IEEE Global Conference for Advancement in Technology (GCAT)

Karnataka, India. Oct 4-6, 2024

A Comprehensive Survey on Sorting Permutations

using Transposition Tree
Ashtapadhi S V∗ , Devi Pillai† , Indulekha T S‡
Department of Computer Science and Applications
Amrita School of Computing
Amrita Vishwa Vidyapeetham, Amritapuri, India
Email: ∗ [email protected], † [email protected], ‡ [email protected]

Abstract—This paper provides a survey on the use of trans- distance between the species. Meanwhile, in computer inter-
position trees for sorting permutations. A permutation can be connection networks, permutations represent nodes, impacting
represented as, π = {n, n − 1, . . . , 1} and its identity as, I = network latency. The interconnection networks are typically
{1, 2, . . . , n − 1, n}. Transforming π to I is termed permuta-
tion sorting and it can be done using various operations. A modelled as graphs. The maximum latency between any pair
transposition tree is one such operation that offers a structured of nodes in computer interconnection networks is determined
approach to permutation sorting, facilitating in-depth analysis by the diameter of the underlying Cayley graph [4]. However,
of permutation structures. A transposition tree T =(VT , ET ) is a finding the diameter of the underlying Cayley graph is not a
spanning tree that gives a higher level of abstraction compared to practically optimal solution. Thus, to efficiently solve these
Cayley graphs. We can place T okens/M arkers on vertices and
each edge is a transposition. Research in sorting permutations problems of finding the distance and the diameter, sorting of
plays a crucial role in simplifying complexities within real-world permutations is done using different operations.
domains, particularly in areas such as the design of computer The operation that we focus on is Transposition trees to
interconnection networks and genomic studies. Effective permu- sort permutations. A transposition tree T comprises n vertices
tation sorting techniques are important for identifying genetic (v1 , v2 ,. . . , vn ) interconnected by edges representing specific
mutations and optimizing network performance. This leads to the
emergence of the problems, of finding the distance between the operations (generators). Each vertex contains a token or
source permutation π and its identity I and finding the diameter marker representing a value. Interchanging tokens between
of the underlying Cayley graph. Due to the complexities of the vertices necessitates a direct edge between them, indicating
underlying Cayley graph, sorting using a transposition tree is the execution of the corresponding operation. Rearranging a
an NP-Hard problem. However, optimal solutions are known for sequence of values (permutation) within this tree is equiv-
certain classes of transposition trees. This paper aims to provide
insights into the known classes of transposition trees such as path, alent to tokenswapping along the connecting paths. Tokens
star, brooms and their corresponding results for the distance can solely move along the branches of the tree. By rearranging
and diameter problems, where available. Additionally, relevant these values within the transposition tree T , it should result
works related to this domain are mentioned to provide a better to a desired or identity configuration with the most less
understanding of advancements in this area. number of swaps possible. The ultimate goal is to reorder the
Index Terms—Sorting, permutation, transposition trees, Cay- original permutation into its sorted arrangement, minimizing
ley graph the number of exchanges required.
Even though sorting permutations using transposition trees
I. I NTRODUCTION is an NP-Hard problem [5], for certain classes of transposition
trees, optimal solutions are already known. For the distance
Gaining proper information regarding the advancement in problem, an optimal algorithm will be developed to sort
the research of sorting permutations using a specific operation permutations using a class of transposition trees. For the
namely the transposition tree, is important. It helps in getting diameter problem, the maximum of the minimum number of
an understanding of this particular domain, the works, and moves required to sort a permutation group which is also
the results contributed by various researchers. These works called the upper bound will be found. Essential details on these
related to the sorting of permutations become beneficial in known optimal solutions, along with specific characteristics
their applications in genomics and computer interconnection and properties of transposition trees that allow for efficient
networks. The significance of these studies in practical appli- sorting algorithms to be applied are provided throughout
cations serves as a strong motivation for researchers to work this survey paper. Most of the relevant existing classes of
deeper into the complexities of permutation sorting. transposition trees and their known results have been provided.
Permutations serve as powerful models for representing These existing optimal and efficient results help to provide a
genomes in genomic studies. The genomic distance between better way for researchers to drive new and more efficient
two species is studied to understand the depth of mutation that results in this domain. Even after the development of some
has occurred over time for the old species to become the new optimal solutions in this domain, there exists a lack of efficient
species [1]–[3]. Thus, it addresses the problem of finding the algorithms for sorting permutations using transposition trees.

979-8-3503-7668-5/24/$31.00 ©2024 IEEE 1

This paves the interest of researchers to explore the research move a block from the beginning of the permutation; block
gaps in this field and come up with efficient solutions. The interchanges, where two blocks of elements swap places in a
methods and approaches taken towards designing efficient given permutation, an approximation algorithm for this oper-
solutions by researchers in their past works can be utilized ation can be found in [9]. If the blocks are adjacent, then one
efficiently during the design of new solutions for the distance obtains the transposition operation; prefix block interchanges,
and diameter problem in the field of sorting permutations using where blocks starting from the beginning is swapped with
the operation of transposition trees. each other; and the transposition trees, which organize and
This survey paper provides a complete overview of the analyze permutations through various types of transpositions.
current development in sorting permutations using transposi- All play a role in shaping the landscape of permutation
tion trees. This includes the known results, existing classes of sorting problems. Using a transposition tree as an operation
transposition trees, classes of transposition trees with research is generally recognized as an NP-Hard problem, indicating
gaps, methods and strategies executed in related works etc. its inherent computational complexity. The markers or tokens
This paper shows the possibility of designing even more new are the values assigned for each vertices and these values can
classes of transposition trees and results. But finding an opti- move along the edges of the transposition tree to attain the
mal solution for every newly designed classes of transposition desired configuration. Here generators are represented as edges
tree may not be practical. Thus approximate results can be of a transposition tree. For specific classes of transposition
found for such classes of transposition trees. Researchers can trees like path, star, and brooms, optimal results are found.
use the best suitable solutions found in various transposition Therefore there is a chance for developing optimal results or
trees discussed in this survey paper for their related works upper bounds if a new class of tree is designed or recognized
as a starting point to develop even better and more efficient whose results are not found yet.
solutions with less work. Since the elements of the permutation group are permu-
tations of a given set, these elements can be used to form
II. BACKGROUND a Cayley graph. A Cayley graph is generated with a set of
The preliminary information utilized in this survey paper vertices and edges. Each vertex in a Cayley graph will be the
is provided in this section. A good background knowledge element of the permutation group. Since the vertices are the
regarding sorting permutations and operations are needed for permutations in a Cayley graph, by applying the generators
deeper understanding of all the concepts mentioned throughout from the permutation group to each vertex, a resultant vertex
this survey. The definitions and outcomes referenced in this is obtained to which an edge can be formed. For a transposition
section can be located in [3], [6]–[8]. tree with n number of vertices, the corresponding Cayley
A permutation is a distinct arrangement of all the elements graph contains n! vertices. Finding the shortest distance be-
in a finite set S = {a1 , a2 , . . . , an }. Each permutation is a tween an initial permutation to a target permutation directly
unique sequence in which each element of the set appears using a Cayley graph or the diameter of the Cayley graph
exactly once. In mathematical terms, a permutation can be becomes impractical since it may take high computational
described as a bijective function that maps the set to itself, time. The study of Cayley graphs provides more understanding
rearranging its elements in a specific order. The number of of the structural properties of these graphs that contribute to
all possible permutations of a set with n elements is given by the general problems of permutation ordering.
n! (n factorial). In mathematical structures, a group < G, ◦ > Efforts to devise optimal algorithms for sorting permutations
consists of a non-empty set G and a binary operation that com- have led to the identification of solutions for specific classes
bines any two elements of G to yield a third element within of transposition trees as mentioned. Those tree structures
G. The fundamental axioms of groups—closure, associativity, exhibit unique properties that can be exploited to enhance the
identity, and invertibility are crucial for understanding their efficiency of sorting algorithms. Understanding the structural
properties. A particular kind of group in which the components intricacies of these tree classes is essential for the development
are permutations of a given set M is called a permutation of algorithms capable of addressing the complexities associ-
group G and its group operation is the composition of these ated with sorting permutations.
permutations. One such example is the symmetric group A subset of the group whose elements, when combined
(Sym(M )), which includes any possible combination of a set through the group operation, can generate all elements of the
and is crucial to understanding symmetry and transformations. group, is called a generating set. Each permutation within
Sorting permutations involves the task of arranging these this generating set is known as a generator. Transforming
ordered sets effectively, a problem with widespread applica- one permutation π1 into another permutation π2 using the
tions in computational fields. There are various operations group operation involves identifying a sequence of generators
that is used to sort permutations, and each method has its ρ1 , ρ2 , . . . , ρk such that π1 ◦ρ1 ◦ρ2 ◦. . .◦ρk = π2 . The smallest
way of measuring how far a given permutation is from k in this sequence represents the shortest distance d(π1 , π2 )
being sorted. Examples of such operations include adjacent from the initial permutation to the target permutation, indicat-
swaps, where two neighbouring elements are exchanged; block ing the minimal sequence of generators required.
transpositions, which involve moving a contiguous block of The fundamental connection between the transformation
elements to a different position; prefix transpositions, which distance and the identity permutation within permutation

2
groups is illustrated by the following mathematical expression. Cayley graph. Since finding the diameter and distance between
By multiplying both sides of the permutation equation by vertices of a Cayley graph is an NP-Hard problem, establishing
the inverse of π2 , denoted as π2−1 , we obtain the equation upper bounds for sorting permutations using transposition trees
π2−1 ◦ π1 ◦ ρ1 ◦ ρ2 ◦ . . . ◦ ρk = π2−1 ◦ π2 = I, where I is is essential for estimating the diameter of the corresponding
the identity permutation. Therefore, d(π1 , π2 ) is equivalent to Cayley graph, which in turn sheds light on the efficiency of
d(π2−1 ◦ π1 , I). sorting algorithms. Without the usage of the Cayley graph,
achieving an optimal outcome solely through the transposition
III. A PPLICATIONS tree proves difficult, leading to the emergence of two distinct
Permutations find diverse applications across computer in- problems in this field.
terconnection networks and genomic studies. It offers signif-
icant benefits in terms of optimizing network performance, A. Two types of problem
understanding evolutionary relationships, and advancing sci- 1) The distance problem: Transforming an initial permuta-
entific knowledge. tion into a desired configuration or its identity while minimiz-
ing computational workload is called the distance problem.
A. Computer interconnection networks There exist methods for determining the distance using the
A computer interconnection network serves to connect de- Cayley graph, but they often demand exponential time, as
vices within a network, which could include processors within when employing algorithms like the Dijkstra′ s algorithm.
a multiprocessor computing system. This network enables the However, these approaches are not practically viable due to
routing of data between processors in such systems, and it is their computational complexity. Thus the goal is to find the
commonly represented as a graph. The performance of this least number of moves to change the input permutation to
network is primarily evaluated based on its diameter. This identity permutation using the transposition tree. To address
diameter represents the maximum distance between any two this issue, the essential input parameters required are the
processors in the network, and it significantly impacts latency. source permutation and the transposition tree.
Cayley graphs, often utilized to model these networks, offer 2) The diameter problem: Finding the upper boundaries
advantages such as sublogarithmic diameters in certain cases, governing the sorting procedures within the domain of transpo-
making them preferable over alternatives like hypercubes [4]. sition trees is called as the diameter problem. In a transposition
tree containing n vertices, there are n! permutations. Sorting
B. Genomic studies each permutation requires a certain number of moves. Thus,
In genomic studies, permutation enables researchers to ana- we identify the maximum among the minimum moves needed
lyze genetic data and infer evolutionary relationships between to sort all permutations within the symmetric group. This value
species. Genomes are often represented as permutations of represents the diameter or upper bound. Even in this problem,
genes, with unsigned permutations used when gene orientation there exist methods for determining the distance using the
isn’t specified. This representation assumes knowledge of gene Cayley graph, but they also often demand exponential time, as
order within each genome, a shared set of genes across all when employing algorithms like BF S. Since these methods
genomes, each gene’s single-copy existence, and each genome are not practically feasible due to their computational de-
being a single chromosome. The distance between species, in mands, it becomes essential to ascertain the upper bound of the
terms of permutations, quantifies the minimal number of swaps transposition tree. The essential requirement for solving this
needed to convert one permutation (representing a species) into problem is having a transposition tree as the input parameter.
another, known as the genomic rearrangement problem. This
distance problem is pivotal in understanding genetic evolution B. Known Results
[1]–[3]. This process aligns with genomic research objectives, The exploration of already existing results related to the
as it enables the study of genetic mutations and evolutionary context of permutations and transposition trees provides us
patterns. The process of transforming a source permutation, with valuable insights. These results contribute to increasing
π into an identity permutation, often denoted as I is known the efficiency of the solutions and help in reducing the
as sorting. Since for every transformation sequence, there complexity of sorting algorithms.
is a corresponding sorting sequence II, we work on sorting 1) Known Results for Diameter Problem :
permutations and thus can use the same generator sequences a) Path: A path is a class of transposition tree with a
used for sorting, to transform the source permutation to a simple linear structure represented as P =(V, E). Path is a
specific target permutation. spanning tree with a set of vertices V = {v1 , v2 , . . . , vn } and
edges E={(v1 , v2 ), (v2 , v3 ) . . . (vn −1, vn )} where its lower
IV. T RANSPOSITION T REES , A R EVIEW bound and the upper bound results are based on the number
Different forms of transposition tree classes have been of inversions [6]. This path can also be represented as (Pn ),
structured by different researchers throughout the years. Those where n is the count of vertices. An inversion in a permutation
classes of transposition trees were studied thoroughly and their occurs when there exists a pair of elements (vi , vj ) where
characteristics were used to develop suitable optimal results i < j and vi > vj . This indicates that each pair of elements
needed. For every transposition tree, there is an underlying that are out of their natural order in the permutation is regarded

3
as an inversion. Optimally, we can sort a path with ′ k ′ swaps vertices of the star part (x) are called star leaves v1 , v2 , . . . , vk ,
where ′ k ′ is the number of unordered pairs or inversions in and the vertices of the path (n − x) are called path vertices
the given permutation. Hence, by examining the inversions vk+1 , . . . , vn where 2 ≤ k ≤ (n − 3). The central vertex,
within a permutation, we can find the minimum number of denoted as vc, serves as the connection point between the path
swaps required to sort it. Thus, the lower bound for sorting a and the star. This center vertex vc, is also a part of the path
permutation using a path is equal to the number of unordered graph. Specifically, vc is one of the path vertices, say vk+1 .
pairs or inversions present in that permutation [6]. Figure 3 shows an example of a Single Broom. A polynomial
time algorithm Ab which runs in O(n2 ) time will efficiently
sort a permutation using a single broom [11].

Fig. 1. path

For example, Figure 1 shows a path with n ver-

tices (n = 5). The inversions within this path are
(4, 2), (4, 1), (4, 3), (2, 1), (5, 1), (5, 3). Consequently, the path Fig. 3. Single Broom
displayed comprises a total of 6 inversions, establishing the
lower bound within this path. b) Double Broom: A double broom is an advanced tree
Upper Bound: The maximum number of inversions in a structure that expands on the concept of the single broom. A
permutation of length n is given by n(n−1) 2 which serves as double broom is constructed by connecting the central vertices
an upper bound on the number of inversions possible [6]. of two stars to both leaf vertices of a path graph. Specifically,
b) Star: A star is a transposition tree where a central the center of one star is connected to one leaf vertex of the
vertex v1 is directly connected to a set of peripheral ver- path, while the center of the other star is connected to the
tices v2 ,v3 ,. . . , vn . Optimal ordering of permutations can be opposite leaf vertex of the path. This structure comprises a
achieved through the use of efficient algorithms that leverage total of n vertices, which are partitioned into left star vertices
star transposition trees. The minimum number of swaps needed v1 , v2 , . . . , vi , path vertices vi+1 , . . . , vj , and right star vertices
to sort tokens initially placed on a star is determined by (u+l) vj+1 , . . . , vn . The centers of both the stars are included as the
where u is the count of unhappy leaves and l is the number path vertices. Figure 4 shows an example of a double-broom
of cycles in the permutation that are at least length 2 and with 11 vertices, where v1 , v2 , v3 represents the left star leaves,
do not include the center vertex [10]. The underlying Cayley v4 , v5 , v6 , v7 , v8 represents the path vertices and v9 , v10 , v11
graph of a star is called a star graph [4]. When it comes to represents the right star leaves. A polynomial time algorithm
swapping pairs of leaf nodes in a star, each swap necessitates 3 Adb which runs in O(n2 ) time will sort a permutation using
moves. Given there are n−1 2 such pairs, the maximum number a double broom ‘ [11].
of movesj required k to sort a permutation with n vertices in a
3(n−1)
star is 2 , establishing an upper bound [4].

Fig. 4. Double Broom

Fig. 2. star These known results contribute foundational benchmarks

for evaluating the efficiency and computational complexities
2) Known Results for Distance Problem: associated with sorting permutations within the context of
a) Single Broom: A Single Broom is a tree structure specific transposition trees. Understanding these bounds and
defined in [11] with n vertices where one of the leaves of complexities aids in comparing and optimizing algorithms for
a path transposition tree is the center vertex of a star. The practical applications.

4
C. Trees with Research gap
There exist some class of transposition trees defined in
certain research papers whose results have not been explored
yet. There is a possibility of finding the upper bound and op-
timal/approximate algorithm to solve the permutation sorting
challenges using these transposition trees.
a) Star Burst Tree: Consider a tree T = (VT , ET ) with
its center denoted as c. Connected to this center are stars
S1 , S2 , . . . , Sb with respective centers c1 , c2 , . . . , cb where
b > 2 via edges (c1 , c), (c2 , c), . . . , (cb , c) [12]. The diameter
of this tree is 4, and it is referred to as a starburstb .
Fig. 7. Millipede tree, M3

D. Implementation
a) Single broom: To sort a permutation efficiently us-
ing a single broom transposition tree, the Ab algorithm is
implemented. Figure 8 shows the implementation of sorting
permutations in a single broom using the Ab algorithm. Here,
the source permutation is sorted to its identity with the
minimum number of adjacent swaps. Let Smin be the star
marker that is closest to its home residing on the path and
Pmax be the largest path marker that is not homed. Then, the
Ab algorithm needs to consider the following 3 steps in the
given order to sort a permutation to its identity using single
broom:
Fig. 5. Starburst, Sb3 1. While (∃ a Smin ), efficiently home Smin ;
2. While (∃ a Pmax ), efficiently home Pmax ;
b) n-broom: A structure known as an n-Broom, where 3. Efficiently home the markers in the star that need to be
n ≥ 3, is created by connecting the terminal nodes of n single homed.
brooms to a shared node [11].

Fig. 6. n − broom, where n = 4

Fig. 8. Single broom implementation
c) Millipede Tree: A millipede tree Mb is formed by
joining the center vertices c1 , c2 , . . . , cb of b path graphs b) Double broom: The implementation of the Adb algo-
p1 , p2 , . . . , pb , where b is an odd number. The count of vertices rithm in a double broom with 9 vertices is shown in figure 9.
in π is established
( by the following function [11]: This shows the efficient sorting of permutations from a source
2i + 1 if 1 ≤ i ≤ ⌈b/2⌉ permutation using double broom. Let SL and SR be the left
n(π) =
2(b − i) + 3 otherwise and right stars respectively, with corresponding center vertices
R L
   of Mb
is (b + 1) and total count of vertices
The diameter CL and CR . Let Smin and Smin be the closest star marker
n = 2× 2b × 2b + 1 −1. Center of the tree vb = c(b+1)/2 residing on the path SR and SL respectively and Pmax be
with eccentricity (b + 1)/2 [11]. the maximum valued unhomed path marker. Then, the Adb

5
algorithm needs to consider the following steps in the given of symmetric graphs that function as processor or computer
order to sort a permutation to its identity using double broom: interconnection networks. These graphs exhibit various prop-
R R
1. While ∃ a Smin marker, efficiently home Smin ; erties that make them particularly useful in this context. One
L L
2. While ∃ a Smin marker, efficiently home Smin ; of the key properties is the diameter of the graph, which
3. (At this point, only the stem is left) While ∃ a Pmax represents the maximum distance between any two vertices.
marker, efficiently home Pmax ; The diameter is crucial for determining the efficiency of
4. Efficiently home the markers of SR and SL . communication within the network. However, calculating the
diameter of a Cayley graph is a complex task with high
computational complexity. Akers and Krishnamurthy [4] were
pioneers in investigating upper bounds for the maximum dis-
tance (diameter) of Cayley graphs derived from transposition
trees. Their research focused on transposition trees, which are
important operations used in finding the diameter or distance
to sort permutations. As a result of their work, certain classes
of transposition trees, such as star and pancake graphs, were
developed. For a star graph with n vertices, referredj to as the
k
n-star graph, the diameter has been found to be 3(n−1) 2 .
This result has significantly contributed to the field of sorting
permutations using transposition trees. Akers and Krishna-
murthy introduced a method based on enumerating permuta-
tions and calculating the maximum number of moves needed
Fig. 9. Double broom implementation
for sorting. The insights gained from studying the diameter of
these graphs have advanced our understanding of sorting and
E. Gaps communication processes within complex network structures.
Various classes of transposition trees are designed and ap- When trying to find the upper bound for sorting permutations
plied to solve the distance and diameter problems. Even more using transposition trees, different techniques can be used. In
classes of transposition trees can be developed and used for [13], Ganeshan proposed an alternative method to compute
resolving the sorting challenges. Based on the current works in an upper bound (β) by analyzing the transposition tree’s
this field, some classes of transposition trees have been already structure. The approach involves deleting pairs of vertices that
designed but remain unexplored without any solutions till now. are diametrically apart. In the paper, Ganeshan focuses on
One such tree is the starburst as a transposition tree defined in assessing existing upper bounds on diameter, proposing a new
[12]. These trees could be studied in detail by examining their O(n2 )algorithm for diameter estimation based on transposi-
characteristics to develop optimal and tight solutions for the tion trees, and analyzing its properties and performance. The
distance and diameter problems in sorting permutations. This algorithm’s effectiveness is demonstrated by proving that its
approach holds the potential to unveil inventive solutions to a computed values are less than or equal to the previously known
problem traditionally viewed as computationally formidable. diameter upper bound.
Also, while the upper bounds for path and star transposition
trees are established, the corresponding upper bounds for Even more algorithms, that compute the upper bound were
sorting permutations using single broom and double broom developed. Chitturi [14] developed and introduced two new
remain unidentified. The already known results can also be algorithms, γ and δ‘. These algorithms were developed to find
utilized effectively while developing an optimal solution for the upper bound on diameter that runs in polynomial time.
sorting permutation using a new class of transposition tree. The γ method focuses on identifying vertices with maximum
By pursuing approximate or optimal algorithms within these distances, known as recluses, and iteratively deleting them
uncharted tree classes, current works not only extend the to calculate the diameter. The δ method is an intermediate
frontiers of algorithmic effectiveness but also pave the way for method that identifies and deletes sets of leaves with maximum
fresh opportunities in real-world applications and advances in eccentricity. A tightest known upper bound(δ*) for a full
computational aspects of permutation sorting. binary tree was then developed surpassing the previously
known best upper bound δ‘ in [12]. This result yields the
V. R ELATED W ORKS smallest cumulative sum among all trees of a given size
Numerous studies within the domain of sorting permutations n ≤ 15. In addition to full binary trees, the paper theoretically
are currently underway, offering valuable insights that can proves that δ* is tighter than δ‘ for several other tree classes.
assist newer research endeavors to handle the challenges It also highlights that the upper bounds computed might
with greater efficiency. Each research study contributes to not be tight for strings, suggesting further exploration in
the collective understanding of the domain, uncovering new this area. It discusses the computation of γ on a double
techniques, algorithms, and insights that enhance knowledge. broom using algorithm Aγ, where careful manipulation of
Cayley graphs play a significant role in the design and analysis incoming markers can improve upon the initial greedy method,

6
indicating possibilities for refining algorithms for certain trees. are established, the corresponding upper bounds for single
A Balanced Starburst is a class of trees which is also defined broom and double broom remain unidentified. The research
in [12]. It is formed when the center of k stars k > 2 has gap identified in this field is that only a few classes of
an edge to a common center vertex (c). A balanced starburst transposition trees are identified and results are known only
guarantees a fairly even spread of vertices across the stars, for a few. A Balanced − Starburst is a class of trees
thereby enhancing the overall symmetry and stability of the which is defined in [12]. It is formed when the center of
tree arrangement. The upper bound or optimal algorithms for k stars k > 2 has an edge to a common center vertex (c).
sorting permutation using Starburst have not been explored Upperbound and optimal algorithms are not identified for
yet. The properties of the structures of other tree classes the starburst tree. The findings of study in this field have
namely, the single broom and double broom were studied the potential to contribute to real-world network optimization
by Indulekha T S and Bhadrachalam Chitturi [11]. They and genomic studies. Furthermore, more classes of trees can
focused on optimal algorithms for sorting permutations using be developed for extending the research in this area. Thus,
these specific tree classes, single-broom and double-broom additional research developments can be made by generating
transposition trees. The study presents efficient algorithms for optimal or approximation algorithms for sorting permutations
sorting permutations with minimal swaps or moves using these using these newly developed classes of trees. By understanding
trees. The research introduces a simple and efficient algorithm the structural intricacies of different tree classes and leveraging
to find the best sequence of swaps for organizing permutations their unique properties, efficient sorting algorithms can be
using the broom transposition tree, as well as an innovative developed to tackle the challenges of permutation ordering.
optimal algorithm for arranging permutations with a double
broom. Moreover, a new category of trees named millipede R EFERENCES
trees is presented, showing that δ* offers a more precise upper
bound for sorting permutations with a balanced millipede tree [1] D. A. Christie, Genome rearrangement problems. University of
Glasgow (United Kingdom), 1998.
compared to δ‘, both of which were previously developed [2] A. R. Oliveira, G. Jean, G. Fertin, K. L. Brito, L. Bulteau, U. Dias,
algorithms. It demonstrates that δ* provides a tighter upper and Z. Dias, “Sorting signed permutations by intergenic reversals,”
bound than δ‘ for a balanced millipede tree. Finally, the IEEE/ACM Transactions on Computational Biology and Bioinformatics,
vol. 18, no. 6, pp. 2870–2876, 2021.
authors also highlight how algorithm Ad b fails to produce an [3] D. Kim, “Sorting on graphs by adjacent swaps using permutation
optimal swap sequence for n-Brooms, underscoring the need groups,” Computer Science Review, vol. 22, pp. 89–105, 09 2016.
for future research to explore polynomial-time algorithms for [4] S. Akers and B. Krishnamurthy, “A group-theoretic model for symmetric
interconnection networks,” IEEE Transactions on Computers, vol. 38,
this structure. no. 4, pp. 555–566, 1989.
The literature on sorting permutations with transposition [5] O. Aichholzer, E. Demaine, M. Korman, A. Lubiw, J. Lynch,
trees encompasses diverse methodologies for estimating up- Z. Masárová, M. Rudoy, V. Williams, and N. Wein, “Hardness of token
per bounds, ranging from enumeration-based approaches to swapping on trees,” in 30th Annual European Symposium on Algorithms,
ESA 2022, ser. Leibniz International Proceedings in Informatics, LIPIcs,
sophisticated algorithms, each contributing to the broader S. Chechik, G. Navarro, E. Rotenberg, and G. Herman, Eds. Germany:
understanding of efficient sorting within permutation groups. Schloss Dagstuhl - Leibniz-Zentrum für Informatik, Sep. 2022, publisher
Diverse research outcomes introduce a variety of methodolo- Copyright: © 2022 Schloss Dagstuhl- Leibniz-Zentrum fur Informatik
GmbH, Dagstuhl Publishing. All rights reserved.; 30th Annual European
gies, approaches, and algorithms. These can be compared, Symposium on Algorithms : ESA 2022, ESA 2022 ; Conference date:
contrasted, and combined to develop more effective and ef- 05-09-2022 Through 09-09-2022.
ficient sorting techniques. Research results cater to the diverse [6] D. E. Knuth, The Art of Computer Programming, Volume 2: Seminu-
merical Algorithms, 3rd ed. Boston: Addison-Wesley, 1997.
needs and requirements of specific applications and scenarios. [7] T. W. Hungerford, “Abstract algebra,” 1990.
Tailoring solutions to different contexts ensures versatility and [8] K. Yamanaka, E. D. Demaine, T. Ito, J. Kawahara, M. Kiyomi,
applicability across a wide range of practical scenarios. [5], Y. Okamoto, T. Saitoh, A. Suzuki, K. Uchizawa, and T. Uno,
“Swapping labeled tokens on graphs,” Theoretical Computer
[8], [12], [14]–[20] have extensively explored and compared Science, vol. 586, pp. 81–94, 2015. [Online]. Available:
different sorting algorithms and their upper bounds for sorting https://www.sciencedirect.com/science/article/pii/S0304397515001656
permutations. [9] J. Pai and B. Chitturi, “Approximation algorithms for sorting permu-
tations by extreme block-interchanges,” Theoretical Computer Science,
VI. C ONCLUSION vol. 891, 09 2021.
[10] A. Biniaz, K. Jain, A. Lubiw, Z. Masárová, T. Miltzow, D. Mondal,
This survey paper focuses on the problem of efficiently A. M. Naredla, J. Tkadlec, and A. Turcotte, “Token swapping on
trees,” Discrete Mathematics &amp; Theoretical Computer Science,
sorting permutations using transposition trees. It explores vol. 24, no 2, no. Discrete Algorithms, Jan. 2023. [Online]. Available:
various tree classes, such as star, path, and brooms, and their http://dx.doi.org/10.46298/dmtcs.8383
key properties. The goal of those trees is to design optimal [11] I. T. Sadanandan and B. Chitturi, “Optimal algorithms for sorting
algorithms that minimize computational workload while trans- permutations with brooms,” Algorithms, vol. 15, no. 7, 2022. [Online].
Available: https://www.mdpi.com/1999-4893/15/7/220
forming an initial permutation into a desired configuration. [12] B. Chitturi and T. Indulekha, “Sorting permutations with a transposition
The research addresses the computational complexity asso- tree,” in 2019 8th International Conference on Modeling Simulation and
ciated with sorting permutations and highlights the need for Applied Optimization (ICMSAO), April 2019, pp. 1–4.
[13] A. Ganesan, “Diameter of cayley graphs generated by transposition
computationally tractable algorithms. While upper bounds for trees,” CoRR, vol. abs/1202.5888, 2012. [Online]. Available:
sorting permutations using path and star transposition trees http://arxiv.org/abs/1202.5888

7
[14] B. Chitturi, “Upper bounds for sorting permutations with a transposition
tree,” Discrete Mathematics, Algorithms and Applications, vol. 05,
no. 01, p. 1350003, 2013.
[15] J. Kawahara, T. Saitoh, and R. Yoshinaka, “The time complexity of
permutation routing via matching, token swapping and a variant,” JGAA,
vol. 23, pp. 29–70, 2019.
[16] D. K. Das, B. Chitturi, and S. S. Kuppili, “An upper bound
for sorting permutations with a transposition tree,” Procedia
Computer Science, vol. 171, pp. 72–80, 2020. [Online]. Available:
https://www.sciencedirect.com/science/article/pii/S1877050920309716
[17] A. K. Suresh, M. Arjun, and T. S. Indulekha, “Analysis of algorithm d*
on n-broom,” in 2023 4th IEEE Global Conference for Advancement in
Technology (GCAT), 2023, pp. 1–6.
[18] B. Chitturi and I. T. S, “Sorting permutations with a transposition
tree,” CoRR, vol. abs/1811.07443, 2018. [Online]. Available:
http://arxiv.org/abs/1811.07443
[19] D. N. Swetha, T. S. Indulekha, and A. S. Nayana, “Sorting permutations
with 3-brooms,” in 2023 4th IEEE Global Conference for Advancement
in Technology (GCAT), 2023, pp. 1–6.
[20] M. J, S. Scaria, and I. T S, “On token swapping in labeled tree,” in 2021
Third International Conference on Inventive Research in Computing
Applications (ICIRCA), 2021, pp. 940–945.