LHC limits on KK-parity non-conservation in the strong sector of universal extra-dimension models

In five-dimensional universal extra-dimensional models compactified on an $S^1/Z_2$ orbifold four-dimensional kinetic terms are allowed at the two fixed points. If these terms are unequal then Kaluza-Klein (KK) parity is broken. Within such a framework we consider resonant production of the n = 1 KK-gluon at the Large Hadron Collider and its subsequent decay to $t\bar{t}$, where both production and decay are KK-parity non-conserving. We use, for the first time, the exclusion data for a $t\bar{t}$ resonance obtained by the ATLAS collaboration to limit the mass ranges of quark and gluon excitations of the KK-parity-violating model which are found to be in the ballpark of 900 GeV.


I Introduction
The detection of a Higgs-like scalar particle at the Large Hadron Collider (LHC) at CERN is a landmark accomplishment. Much activity is now aimed to uncover detailed properties of this new state and compare it with the expectations from the standard model (SM). There is also a continuing interest regarding the physics which lies beyond the standard model. The evidence for such physics, albeit indirect, can be traced to the issue of naturalness of the Higgs-scalar mass, the observed masses and mixing of neutrinos, and the quest for a dark matter candidate. The energy scale for new physics remains unknown but there are several motivations which encourage us to expect that it may well be within the reach of the LHC. Here our intention is to constrain a class of non-minimal universal extra-dimensional models where the lowest (n = 1) Kaluza-Klein excitations are not stabilized by any symmetry. We show that the data reported by the ATLAS [1,2] and the CMS [3,4] Collaborations excluding a heavy resonance in the tt channel eliminate significant regions in the n = 1 KK-gluon and KK-quark mass plane.
In the simplest Universal Extra Dimension (UED) model there is one extra flat spacelike dimension and it is accessible to all particles [5]. The extra dimension y is compact (radius of compatification R) and has a Z 2 symmetry (y → −y) to incorporate chiral fermions. For every SM particle one has a tower of KK excitations, each member being specified by an integer n = 0, 1, 2, . . ., the standard model particle being the n = 0 mode of the tower. The SM masses of the particles are small compared to 1/R and it is a good approximation to take the KK states for all particles at any level n to be degenerate with a mass n/R.
The Z 2 symmetry produces a conserved KK-parity which is (−1) n for the n-th KK-level. The SM particles are of even parity while those of the first level are odd. KK-parity ensures that the lightest among the n = 1 particles is absolutely stable and so can be a dark matter candidate, the Lightest Kaluza-Klein Particle (LKP). This is the essence of the Universal Extra Dimension (UED) Model.
The above S 1 /Z 2 orbifold compactification has two fixed points at y = 0 and y = πR. At these boundary points inclusion of additional four-dimensional interactions is allowed by the symmetry. In fact, these terms are useful as counterterms for compensating loop-induced contributions [6] of the five-dimensional theory. In the simplest choice, the minimal Universal Extra-Dimensional Models (mUED) [7,8], these terms are chosen so that they exactly cancel the five-dimensional loop effects at the cutoff scale of the theory Λ. Thus these boundary contributions, e.g., logarithmic corrections to masses of KK particles, are such that they vanish at the scale Λ. At lower energies these contributions remove the mass degeneracy among states at the same KK-level n.
The radius of compactification R sets the mass scale of the theory and splittings within the KK-states of the same level are controlled by the cutoff Λ. They can be constrained by using known measurement results. Thus, from the muon (g − 2) [9], flavour changing neutral currents [10,11,12], Z → bb decay [13], the ρ parameter [5,14], and other electroweak precision tests [15,16], one has typically 1/R ∼ > 300 − 600 GeV. Comparing the production and leptonic decay of n = 2 electroweak gauge bosons with the CMS LHC data a limit of 1/R ∼ > 715 GeV has been placed [17].
In this work we explore the scenario where the boundary terms depart from their special choice of mUED. In addition, the boundary terms at the two fixed points are allowed to be unequal. This leads to non-conservation of KK-parity and opens the door for n = 1 KK-states to be produced singly in SM particle collisions and also for them to decay to n = 0 states 1 some of which has been emphasized in an earlier work [18]. Because of these features the picture considered here is termed KK-parity violating UED.
Here we examine the coupling of the n = 1 KK-gluon to a pair of SM-quarks (n = 0 states). Such couplings are absent in mUED and are hallmarks of the non-minimality discussed above. They provide an avenue for comparing the predictions of the theory with measurements at the LHC.
Specifically, we consider resonant production of the n = 1 KK-gluon in pp collisions at the LHC through the KK-parity non-conserving coupling. We take the KK-gluon to be the lightest n = 1 level particle. Once produced, KK-conserving decays being kinematically disallowed, the KK-gluon decays to a pair of zero-mode quarks through the same KK-parity-violating coupling. The branching ratio is equal for all types of quarks and hence it is 1/6 for the tt mode which we examine. Both the ATLAS and CMS collaborations have looked for the signal of a resonance produced in pp collisions which decays to a pair of top-antitop quarks. Results from ATLAS for the 7 TeV [1] and 8 TeV [2] LHC runs are now available as are the corresponding findings of CMS in [3] and [4] respectively. From the lack of observation of such a state, 95% C.L. upper limits on the cross section times branching ratio of such a signal as a function of the resonance mass have been reported in these publications. Here we use the limits from the 8 TeV runs to constrain the masses of the n = 1 level KK quarks and gluons of the model. We would like to emphasize that this is the first effort to restrict the parameters of KK-parity-violating UED using existing LHC data.
The two essential ingredients for calculating the signal are the mass of the n = 1 gluon state and the strength of its KK-parity-violating couplings. In the following section we briefly review the UED scenario with boundary-localized kinetic terms and lead up to the KK-excitation masses in such a framework. In the next section we calculate the Z 2 -parity-violating coupling involving the first excitation of the KK-gluon and a zero-mode quark-antiquark pair. With these results we then derive the expected tt signal from the production of the KK-gluon at the LHC and its subsequent decay. This is compared with the CMS [4] and ATLAS [2] 8 TeV results and the restrictions on the KK-excitation masses are exhibited. Our conclusions appear at the end.

II KK-parity-violating UED, KK masses
In nonminimal UED one can consider kinetic and mass terms localized at the fixed points. Here we restrict ourselves to boundary-localized kinetic terms only [19] - [24], [18]. Thus we consider a five-dimensional theory with additional four-dimensional kinetic terms at the boundaries at y = 0 and y = πR.
We illustrate the idea by considering free fermion fields Ψ L,R whose zero modes are the chiral projections of the SM fermions. The five-dimensional action with BLKT is [25] Here σ µ ≡ (I, σ) andσ µ ≡ (I, − σ), σ being the (2 × 2) Pauli matrices. r a f , r b f are the strengths of the boundary terms which are chosen to be the same for Ψ L and Ψ R for simplicity.
It is helpful to express five-dimensional fermion fields using two component chiral spinors 2 [25]: Below we examine the case of Ψ L in detail. The results for Ψ R will be similar and can be obtained by making appropriate changes.
Variation of the action functional Eq. (1) utilising Eq.
(2) results in coupled equations for the y-dependent wave-functions, f n L , g n L . Following routine steps 3 one finds: 1 + r a f δ(y) + r b f δ(y − πR) m n f n L − ∂ y g n L = 0, m n g n L + ∂ y f n L = 0, (n = 0, 1, 2, . . .). (4) Eliminating g n L one obtains the equations: From now onwards we drop the subscript L on the wave-functions and denote them simply by f and g.
The boundary conditions are [20] f n (y) One then obtains the solutions: f n (y) = N n cos(m n y) − r a f m n 2 sin(m n y) , 0 ≤ y < πR, f n (y) = N n cos(m n y) + r a f m n 2 sin(m n y) , −πR ≤ y < 0.
where the masses m n for n = 0, 1, . . . are obtained from the transcendental equation [20]: The solutions are orthonormal, i.e.: These wave-functions are combinations of a sine and a cosine function unlike in the case of mUED where they are one or the other of these two trigonometric functions. This difference and that the KK masses are solutions of Eq. (9) rather than just n/R are at the root of the distinguishing features of this model.
In this paper we examine two versions of KK-parity-violating UED. In one we take symmetric boundary-localized terms for fermions, i.e., r a f = r b f ≡ r f . The other case has the BLKT at one of the fixed points only: r a f = 0, r b f = 0. In this second alternative Eq. (9) becomes The normalisation constant N n is determined from orthonormality. When r a For the other case when r b f = 0 we use r a f ≡ r f and one has In this work we deal only with the zero modes and the n = 1 KK wave-functions of the five-dimensional fermion fields.
The action for the five-dimensional gluon field, G N (N = 0 . . . 4) with BLKT r a g , r b g at the fixed points can be similarly written down. It is straightforward to show following similar steps 4 that in the G 4 = 0 gauge the gluon field has the KK-expansion: where the functions a n (y) are of the same form as Eq. (8). In this case the five-dimensional contributions to the KK-gluon mass, m n , satisfy which is identical to Eq. (9) for fermions.
For the other case that we also consider (r a g = 0, r b g = 0) the transcendental equation (15) reduces to tan(m n πR) = − r a g m n 2 .
This equation is the same as Eq. (11) for fermions with similar BLKT.
As the KK-masses of fermions and gauge bosons are obtained from similar equations it is convenient to discuss the solutions together. Below we use r a α , r b α to stand for the BLKT strengths with α = f or g. Our focus will be on the n = 1 state.
as a function of R α when the BLKT is present only at the y = 0 fixed point. The inset shows a blow-up of the region of R α that is considered later. Note the very mild variation of M (1) in the insets of both panels. The plots apply for α = f (fermions) and g (gluons).
In Fig. 1 we show the variation of M (1) ≡ m α (1) R, which is dimensionless, in the two alternate cases. In the left panel M (1) is shown as a function of R α ≡ r a α /R in the symmetric (r a α = r b α ) limit. In the inset is presented the dependence of M (1) on the asymmetry parameter ∆R α , for several choices of R α . The range of variation of ∆R α shown in the inset is what we use in our later discussion. It is seen that M (1) changes very slowly. Note also that the mass of the n = 1 state for a particular R α always remains more than that corresponding to any larger R α for the entire variation of ∆R α . Therefore, irrespective of the value of ∆R α , the mass ordering within the n = 1 level is determined on the basis of R α . In the right panel, we present M (1) as a function of R α = r a α /R when the BLKT is present only at y = 0. In the inset is shown the region in which we later choose the gauge boson BLKT. Here too the variation of M (1) is hardly significant. The key message from both panels is that the KK-mass falls with increasing R α ; the fermion or gauge boson with the largest R α is the lightest n = 1 KK state. In this work we keep R g > R f to ensure that the n = 1 gluon is lighter than the quarks of the same level. So, the former cannot decay via KK-number conserving modes (which would have dominated if allowed) and the branching ratio to tt is 1/6.

III Coupling of G 1 with zero-mode quarks
Besides the masses, the other ingredient required for the proposed calculation is the strength of the G 1 f 0 f 0 coupling. It is given by The five-dimensional gauge coupling g 5 which appears above is related to the usual coupling g through As noted in the previous section, the wave-functions f 0 L , g 0 R for zero-mode quarks and a 1 for the KK-gluon depend on the choices made for the boundary localized terms.
We remark in passing that, irrespective of the nature of the gluon boundary terms, the coupling g G 1 G 0 G 0 is always zero. Thus the resonant production of G 1 is initiated only by quarks and antiquarks in the colliding particles and the gluonic content of the proton plays no role.

III.1 BLKT at both fixed points
The first option which we study has BLKTs of the same strength at the two fixed points for quarks (r a f = r b f = r f ) while for the gauge bosons (r a g = r b g ). The y-dependent wave-functions of our interest here are found to be and with where we have used as earlier M (1) ≡ m g (1) R, and the scaled dimensionless variables Using the above we get In the left panel of Fig. 2 we plot the square of the coupling for a fixed value 5 of R a g = 3.0 as a function of R f for several values of ∆R g . It is seen that the strength of the coupling decreases as R f increases while it increases as ∆R g increases.
Physics consequences of these couplings are discussed in the next section. At this stage we urge the reader to note that the KK-parity-violating coupling gets smaller as R f tends towards R g i.e., as the splitting among the n = 1 KK-excitations is decreased. Also, it can be readily seen using Eq. (15) that if R a g = R b g , i.e., the BLKTs are symmetric at y = 0 and y = πR for the gauge boson, as chosen for the quarks, the coupling in Eq. (24) vanishes. This can be traced to a y ←→ (y − πR) Z 2 -symmetry of the theory for this choice which forbids an n = 1 state to couple exclusively to zero modes. In general, g G 1 f 0 f 0 decreases as ∆R g gets smaller.

III.2 BLKT at one fixed point
In the second case, for both the quarks and the gauge bosons we assume that the BLKT are present at only the y = 0 fixed point. The y-dependent wave-functions in this case are and where R g ≡ r g /R. With these wave-functions we get for this case In the right panel of Fig. 2 we plot the square of the coupling strength as a function of R f for several choices of R g for this alternative. In order to keep the n = 1 KK-gluon lighter than the quarks of the same level we have kept R g > R f . It is to be noted that the strength of the coupling decreases as R f increases but it increases as R g increases. The coupling vanishes if R g = R f , as can be seen from Eq. (27).
IV Single G 1 production and decay to tt From here onwards we will not explicitly write the KK-number (n = 0) as a superscript for the SM particles. Wave-functions with no superscripts are for the SM particles.   We now investigate the single production of G 1 at the LHC. We consider the process pp (qq) → G 1 followed by G 1 → tt. Note that both the production and decay of G 1 involves the KKparity-violating coupling. The signature of this mode would be a resonance in the tt channel at the G 1 mass. The ATLAS and CMS collaboration have both searched for such a resonance in pp collisions at 7 and 8 TeV [1, 2, 3, 4]. From non-observation of such a signal 95% C.L. upper bounds have been placed on the cross section times branching ratio as a function of the mass of a tt resonance. Comparing these bounds with the calculated values in the KK-pariy-violating framework enables the restriction of the parameter space of the model. To get the most up-todate bounds we use the latest 8 TeV results. At this energy CMS has published [4] the analysis of 19.7 f b −1 of data while ATLAS has presented [2] bounds from 14.3 f b −1 of data. We use the former in our considerations below but also remark on the constraint following from the latter.
The key quantities here are the KK-gluon mass and its coupling to zero-mode quarks. In nonminimal UED, the mass of G 1 is determined from Eqs. (15) and (16) by 1/R and the gluon BLKT R a,b g . The resonance masses excluded by the ATLAS and CMS results are bounds on the n = 1 gluon mass in this model. This restricts 1/R and R a,b g . Further, the single production and the decay of n = 1 KK-excitations of gluons to SM quarks are driven by KK-parity-violating couplings which depend on the quark BLKT R f and also vanish unless the strengths of the gauge BLKT parameters localized at the two fixed points are different, i.e., ∆R g = 0. A specific upper bound on the event rate as quoted by CMS [4] therefore translates to constraints on the above parameters and thence to the mass of the KK-excitations of quarks.
As noted earlier, in spite of the onset of KK-parity violation the coupling g G 1 GG vanishes identically. Consequently, the production of G 1 in pp collisions is driven solely by qq fusion. An analytic expression for the resonant production cross section of the n = 1 KK-gluon from qq fusion in the collision of two protons can be expressed in a compact form: (1) ) + q i ↔q i (28) Here, q i andq i stand for a generic quark and the corresponding antiquark of the i-th flavour respectively. f q i p (fq i p ) is the quark (antiquark) distribution function within a proton. τ ≡ m 2 g (1) /S P P , where √ S P P is the proton-proton centre of momentum energy. Γ(G 1 → q iqi ) represents the decay width of G 1 into the quark-antiquark pair and is given by Here g G 1 qq is the KK-parity-violating coupling of the n = 1 gluon with the SM quarks -see Eqs. (24) and (27).
Eq. (28) represents the lowest order result in QCD. We have not considered higher order contributions in our analysis and used it bearing in mind that QCD corrections usually enhance cross sections and so our results are probably conservative.
To obtain the numerical values of the cross sections, we use a parton-level Monte Carlo code with parton distribution functions as parametrized in CTEQ6L [26]. We take the pp centre of momentum energy to be 8 TeV. Renormalisation (for α s ) and factorisation scales (in the parton distributions) are both set at m g (1) .
We are now ready to present the results of our investigation. To our knowledge, this is the first of its kind where experimental data from the LHC have been used to restrict the parameter space of KK-parity-violating UED. Results for two distinct cases, either BLKTs are present at both fixed points or only at one of the two, will be presented in following two subsections.

IV.1 BLKT at both fixed points
In Fig. 3 we show the region of parameter space excluded by the CMS 8 TeV data [4] for the case in which the quark BLKTs are equal at the two fixed points but KK-parity is broken by the unequal values of the gauge BLKTs. The three panels correspond to different choices of R a g . In each panel the region to the left of a curve in the m g (1) − R f plane is disfavoured by the CMS data. The curves in any one panel correspond to different choices of ∆R g as indicated.
Since the KK-mass is rather insensitive to ∆R g , for a chosen R g there is one-to-one correspondence of m g (1) with 1/R which is shown on the upper axis of the panels. Also, R f determines M f (1) = m f (1) R and is displayed on the right-side axis from where m f (1) corresponding to any 1/R can be read off.
The results depicted in Fig. 3 can be readily understood by noting that the LHC exclusion plots translate in this model to a limit on the KK-parity violating coupling for any chosen n = 1 gluon mass. R a g is fixed for a panel. In any panel, the 95% C.L. CMS upper limit on the cross section times the branching ratio implies that the points of intersection of the curves with a vertical line corresponding to a fixed m g (1) identify (R f , ∆R g ) pairs which lead to the same magnitude of the coupling constant. From the left panel of Fig. 2 it is seen that this happens if an increase in R f is matched by an increase in ∆R g . This is indeed seen to be the case. One can also easily explain the nature of the plots from the standpoint of a fixed R f . As one moves from left to right keeping R f fixed, i.e., to increasing m g (1) , the production cross section falls. It can be checked that though the CMS data also decreases with increasing resonance mass this is slower than the kinematic reduction. Therefore in order to match the observed results a larger KK-violating coupling is needed as m g (1) increases. It is seen from the left panel of Fig. 2 that for a fixed R f the coupling grows as ∆R g is increased. This feature agrees with the results in Fig. 3. R f M f (1) m g (1) (TeV)  The implications of the above results on the n = 1 level KK mass spectrum can be extracted by considering them in conjunction with Figs. 1 and 2. The limits on m g (1) are essentially those on the tt resonance given in the data [4]. However, for any chosen R a g (any one panel) this entire range cannot be covered. This is because the KK-parity violating coupling, which serves to match the model prediction for the cross section for a particular m g (1) with the data, varies only over a limited range (see left panel of Fig. 2). This therefore determines the KK-gluon mass band permissible for a certain R a g . At the same time this puts an upper bound on the n = 1 quark mass, which anyway has to be heavier than the n = 1 gluon in this model. Thus the quark KK excitation mass has to be in a limited range to agree with the LHC data. This feature can be illustrated by a few examples from Fig. 3. From the left panel (R a g = 12.0) one finds that if m g (1) = 625 GeV then m q (1) is bounded from above by 1.77 TeV. If, on the other hand m g (1) = 1.60 TeV the n = 1 quark is constrained (see the right panel, R a g = 3.0) to lie between 1.60 and 2.32 TeV. In Fig. 3 the three R a g choices cover the entire CMS exclusion range of tt resonance mass.

IV.2 BLKT at one fixed point
Now let us turn to the case of quark and gauge BLKTs at only one fixed point. In the left (right) panel of Fig. 4 we show the bounds obtained using the 8 TeV results of CMS (ATLAS). The relevant KK-parity-violating couplings vanish when 6 R f = R g . We show in this case the exclusion curves in the m g (1) − R f plane for different choices of R g . The region below a curve has been ruled out from the data. As expected the CMS bounds based on 19.7 f b −1 data are more restrictive than those from the 14. Our discussion in the following is based on the left panel of Fig. 4. The nature of behaviour seen in Fig. 4 tallies with our earlier considerations. As we have noted in the right panel of Fig. 1, M (1) ≡ m g (1) R is quite insensitive to the value of R g . It is a good approximation therefore to take the mass of G 1 to be simply proportional to 1/R; the 1/R values are indicated in the upper axes of the panels in Fig. 4. For any m g (1) the CMS data gives a bound for the corresponding cross section times branching ratio. Once the mass is fixed, the experimental limit can be achieved by a specific value for the KK-number violating coupling, i.e., the points of intersection of a vertical line with the curves give (R g , R f ) pairs which result in this fixed coupling. This can be borne out by comparing with the right panel of Fig. 2. A second option is to consider the plots in Fig. 4 for a fixed R f . In this case as m g (1) increases the enhanced mass hinders the production of the n = 1 KK gluon. We have checked that the fall seen in the CMS data with increasing resonance mass is slower then this kinematic reduction. To compensate for this, the KK-violating coupling must increase as we move to larger m g (1) . As seen from the right panel of Fig. 2, for a fixed R f the coupling enhancement is accomplished by increasing R g . This is the case in Fig. 4.
The conclusions that can be drawn from the above results are similar to the ones from Fig.  3

V Conclusions
Universal extra dimension models are among the attractive options for beyond the standard model physics. Here the SM particles are complemented with heavier KK excitations which are equispaced in mass. The interaction strengths of these states are controlled entirely by the SM.
Various aspects of the model ranging from constraints from precision measurements to collider searches have been looked at in the literature. Signals for UED are being actively searched for at the LHC.
One of the less attractive predictions of UED is the compressed mass spectrum of KK excitations of all SM particles at any fixed level. A remedy for this had been noted early on. It was shown [7] that five-dimensional radiative corrections split the degeneracy in a calculable way determined by the SM charges of the zero-mode states. The corrections are encoded as additional fourdimensional interactions located at the the two boundary points (BLTs). In this version of UED, known as minimal UED, the practice has been to assume that the couplings of the KK excitations continue to be as for the SM particles and only the mass degeneracy is removed.
In this work we examine departures of the boundary localized kinetic terms from the above minimal choice. There are two possibilities of choosing the four-dimensional kinetic terms at the fixed points with rather distinct physics consequences. In the first, the BLKTs are of equal strength at both fixed points (y = 0, πR). Here, a Z 2 symmetry y ←→ (y − πR) survives. One ends up with a theory where the spectrum of KK-particles and the couplings can be drastically different from mUED. The lightest among the n = 1 KK particles is stable and can be a dark matter candidate [27,28]. The other alternative is to allow the BLKTs at the two fixed points to be of unequal strengths. This will lead to a breakdown of KK-parity and will allow, for example, single production of n = 1 KK-excitations and their decay to SM particles. Earlier we have examined, B 1 (W 1 3 ) → e + e − , µ + µ − , decays after the production of the B 1 (W 1 3 ) singly at the LHC [18].
In this article, we have considered the possible BLKTs for an interacting theory of quarks and gluons. In one alternative, the strengths of quark BLKTs at the two fixed points have been assumed to be equal ≡ r f . For the gauge boson boundary kinetic terms we have considered the general case of unequal BLKTs (r a g = r b g ). Equality of the latter strengths would restore a Z 2parity. As an alternate possibility we have considered the situation where the quark and gluon BLKTs are present only at the y = 0 fixed point. In both cases the boundary terms modify the field equations in the y-direction. Consistency conditions of the solutions of the above equations lead to the masses of KK-excitations of quarks and the gluons and their wave-functions in the y-direction.
We have calculated the coupling of G 1 , the n = 1 KK-excitation of the gluon, to a pair of zero-mode quarks (i.e., SM quarks) as a function, in the first alternative, of r f , r a g , r b g and 1/R. In general, we have presented the coupling as a function of the scaled variable R f for several choices of the other parameters. A similar KK-parity-violating coupling, which arises when the BLKTs are present only at y = 0, has also been evaluated. The coupling is a hallmark of KK-parity violation and vanishes in the ∆R g = 0 limit in the first case and for R f = R g in the second.
These results are utilized to calculate the production of G 1 singly at the LHC and its subsequent decay to tt, both production and decay being via the KK-parity-violating coupling. The predictions are compared with the results on tt resonance production signature at the LHC running at 8 TeV pp centre of momentum energy [2, 4]. It is revealed that nonobservation of this signal with 19.7 f b −1 accumulated luminosity already disfavors a large part of the parameter space (spanned by r f , r a g , r b g and 1/R in one case and r f , r g and 1/R in the other). In the models considered here the n = 1 gluon is lighter than the corresponding quark and the bounds on the mass of the former are the same as that on the tt resonance from the data. The cross section limits from LHC put tight upper bounds on the n = 1 quark excitation mass. In particular, while a range of a few hundred GeV is still permitted for this mass in the first scenario, in the second the n = 1 quarks and gluons have to be quasi-degenerate.
A similar new physics signal can also arise from other models, e.g., if there are extra Z-like bosons as in the Left-Right symmetric models or those with an extra U(1) symmetry. Here we have not attempted to compare the KK-parity-violating UED signals with those from such other scenarios.