Cross-power Spectrum Research Articles

Localizing speakers in multiple rooms by using Deep Neural Networks

In the field of human speech capturing systems, a fundamental role is played by the source localization algorithms. In this paper a Speaker Localization algorithm (SLOC) based on Deep Neural Networks (DNN) is evaluated and compared with state-of-the art approaches. The speaker position in the room under analysis is directly determined by the DNN, leading the proposed algorithm to be fully data-driven. Two different neural network architectures are investigated: the Multi Layer Perceptron (MLP) and Convolutional Neural Networks (CNN). GCC-PHAT (Generalized Cross Correlation-PHAse Transform) Patterns, computed from the audio signals captured by the microphone are used as input features for the DNN. In particular, a multi-room case study is dealt with, where the acoustic scene of each room is influenced by sounds emitted in the other rooms. The algorithm is tested by means of the home recorded DIRHA dataset, characterized by multiple wall and ceiling microphone signals for each room. In detail, the focus goes to speaker localization task in two distinct neighboring rooms.As term of comparison, two algorithms proposed in literature for the addressed applicative context are evaluated, the Crosspower Spectrum Phase Speaker Localization (CSP-SLOC) and the Steered Response Power using the Phase Transform speaker localization (SRP-SLOC). Besides providing an extensive analysis of the proposed method, the article shows how DNN-based algorithm significantly outperforms the state-of-the-art approaches evaluated on the DIRHA dataset, providing an average localization error, expressed in terms of Root Mean Square Error (RMSE), equal to 324 mm and 367 mm, respectively, for the Simulated and the Real subsets.

Upgrade of the Mirnov probe arrays on the J-TEXT tokamak.

The magnetic diagnostic of Mirnov probe arrays has been upgraded on the J-TEXT tokamak to measure the magnetohydrodynamic instabilities with higher spatial resolution and better amplitude-frequency characteristics. The upgraded Mirnov probe array contains one poloidal array with 48 probe modules and two toroidal arrays with 25 probe modules. Each probe module contains two probes which measure both the poloidal and the radial magnetic fields (Bp and Br). To ensure that the Mirnov probe possess better amplitude-frequency characteristics, a novel kind of Mirnov probe made of low temperature co-fired ceramics is utilized. The parameters and frequency response of the probe are measured and can meet the experiment requirement. The new Mirnov arrays have been normally applied for a round of experiments, including the observation of tearing modes and their coupling as well as high frequency magnetic perturbation due to the Alfvén eigenmode. In order to extract useful information from raw signals, visualization processing methods based on singular value decomposition and cross-power spectrum are applied to decompose the coupled modes and to determine the mode number.

Foreground and Sensitivity Analysis for Broadband (2D) 21 cm–Lyα and 21 cm–Hα Correlation Experiments Probing the Epoch of Reionization

Abstract A detection of the predicted anticorrelation between 21 cm and either Lyα or Hα from the epoch of reionization (EOR) would be a powerful probe of the first galaxies. While 3D intensity maps isolate foregrounds in low- k ∥ modes, infrared surveys cannot yet match the field of view and redshift resolution of radio intensity mapping experiments. In contrast, 2D (i.e., broadband) infrared intensity maps can be measured with current experiments and are limited by foregrounds instead of photon or thermal noise. We show that 2D experiments can measure most of the 3D fluctuation power at k < 0.2 Mpc−1 while preserving its correlation properties. However, we show that foregrounds pose two challenges: (1) simple geometric effects produce percent-level correlations between radio and infrared fluxes, even if their luminosities are uncorrelated; and (2) radio and infrared foreground residuals contribute sample variance noise to the cross spectrum. The first challenge demands better foreground masking and subtraction, while the second demands large fields of view to average away uncorrelated radio and infrared power. Using radio observations from the Murchison Widefield Array and near-infrared observations from the Asteroid Terrestrial-impact Last Alert System, we set an upper limit on residual foregrounds of the 21 cm–Lyα cross-power spectrum at z ∼ 7 of Δ 2 < 181 ( kJy sr − 1 mK ) (95%) at ℓ ∼ 800 . We predict levels of foreground correlation and sample variance noise in future experiments, showing that higher-resolution surveys such as LOFAR, SKA-LOW, and the Dark Energy Survey can start to probe models of the 21 cm–Lyα EOR cross spectrum.

Halo Pressure Profile through the Skew Cross-power Spectrum of the Sunyaev–Zel’dovich Effect and CMB Lensing in Planck

Abstract We measure the cosmic microwave background (CMB) skewness power spectrum in Planck, using frequency maps of the HFI instrument and the Sunyaev–Zel’dovich (SZ) component map. The two-to-one skewness power spectrum measures the cross-correlation between CMB lensing and the thermal SZ effect. We also directly measure the same cross-correlation using the Planck CMB lensing map and the SZ map and compare it to the cross-correlation derived from the skewness power spectrum. We model fit the SZ power spectrum and CMB lensing–SZ cross-power spectrum via the skewness power spectrum to constrain the gas pressure profile of dark matter halos. The gas pressure profile is compared to existing measurements in the literature including a direct estimate based on the stacking of SZ clusters in Planck.

Foreground mitigation strategy for measuring the 21 cm-LAE cross-correlation

AbstractThe cross power spectrum of the 21 cm signal and Lyman-α emitters (LAEs) is a probe of the Epoch of Reionization. Astrophysical foregrounds do not correlate with the LAE distribution, though the foregrounds contribute to the error. To study the impact of foregrounds on the measurement, we assume realistic observation by the Murchison Widefield Array using a catalogue of radio galaxies, a LAE survey by the Subaru Hyper Supreme-Cam and the redshift of LAEs is determined by the Prime Focus Spectrograph. The HI distribution is estimated from a radiative transfer simulation with models based on results of radiation hydrodynamics simulation. Using these models, we found that the error of cross power spectrum is dominated by foreground terms. Furthermore, we estimate the effects of foreground removal, and find 99% of the foreground removal is required to detect the 21 cm-LAE signal at k ∼ 0.4 h Mpc−1.

Probing Large-scale Coherence between Spitzer IR and Chandra X-Ray Source-subtracted Cosmic Backgrounds

Abstract We present new measurements of the large-scale clustering component of the cross-power spectra of the source-subtracted Spitzer-IRAC cosmic infrared background and Chandra-ACIS cosmic X-ray background surface brightness fluctuations Our investigation uses data from the Chandra Deep Field South, Hubble Deep Field North, Extended Groth Strip/AEGIS field, and UDS/SXDF surveys, comprising 1160 Spitzer hours and ∼12 Ms of Chandra data collected over a total area of 0.3 deg2. We report the first (>5σ) detection of a cross-power signal on large angular scales >20″ between [0.5–2] keV and the 3.6 and 4.5 μm bands, at ∼5σ and 6.3σ significance, respectively. The correlation with harder X-ray bands is marginally significant. Comparing the new observations with existing models for the contribution of the known unmasked source population at z < 7, we find an excess of about an order of magnitude at 5σ confidence. We discuss possible interpretations for the origin of this excess in terms of the contribution from accreting early black holes (BHs), including both direct collapse BHs and primordial BHs, as well as from scattering in the interstellar medium and intra-halo light.

Determining the H i content of galaxies via intensity mapping cross-correlations

We propose an innovative method for measuring the neutral hydrogen (HI) content of an optically-selected spectroscopic sample of galaxies through cross-correlation with HI intensity mapping measurements. We show that the HI-galaxy cross-power spectrum contains an additive shot noise term which scales with the average HI brightness temperature of the optically-selected galaxies, allowing constraints to be placed on the average HI mass per galaxy. This approach can estimate the HI content of populations too faint to directly observe through their 21cm emission over a wide range of redshifts. This cross-correlation, as a function of optical luminosity or colour, can be used to derive HI-scaling relations. We demonstrate that this signal will be detectable by cross-correlating upcoming Australian SKA Pathfinder (ASKAP) observations with existing optically-selected samples. We also use semi-analytic simulations to verify that the HI mass can be successfully recovered by our technique in the range M_HI > 10^8 M_solar, in a manner independent of the underlying power spectrum shape. We conclude that this method is a powerful tool to study galaxy evolution, which only requires a single intensity mapping dataset to infer complementary HI gas information from existing optical and infra-red observations.

Cross-correlation of galaxies and galaxy clusters in the Sloan Digital Sky Survey and the importance of non-Poissonian shot noise

We present measurements of angular cross power spectra between galaxies and optically-selected galaxy clusters in the final photometric sample of the Sloan Digital Sky Survey (SDSS). We measure the auto- and cross-correlations between galaxy and cluster samples, from which we extract the effective biases and study the shot noise properties. We model the non-Poissonian shot noise by introducing an effective number density of tracers and fit for this quantity. We find that we can only describe the cross-correlation of galaxies and galaxy clusters, as well as the auto-correlation of galaxy clusters, on the relevant scales using a non-Poissonian shot noise contribution. The values of effective bias we finally measure for a volume-limited sample are $b_{cc}=4.09 \pm 0.47$ for the cluster auto-correlation and $b_{gc}=2.15 \pm 0.09$ for the galaxy-cluster cross-correlation. We find that these results are consistent with expectations from the auto-correlations of galaxies and clusters and are in good agreement with previous studies. The main result is two-fold: firstly we provide a measurement of the cross-correlation of galaxies and clusters, which can be used for further cosmological analysis, and secondly we describe an effective treatment of the shot noise.

Measurement of acoustic velocity components in a turbulent flow using LDV and high-repetition rate PIV

The present study provides theoretical details and experimental validation results to the approach proposed by Minotti et al. (Aerosp Sci Technol 12(5):398–407, 2008) for measuring amplitudes and phases of acoustic velocity components (AVC) that are waveform parameters of each component of velocity induced by an acoustic wave, in fully turbulent duct flows carrying multi-tone acoustic waves. Theoretical results support that the turbulence rejection method proposed, based on the estimation of cross power spectra between velocity measurements and a reference signal such as a wall pressure measurement, provides asymptotically efficient estimators with respect to the number of samples. Furthermore, it is shown that the estimator uncertainties can be simply estimated, accounting for the characteristics of the measured flow turbulence spectra. Two laser-based measurement campaigns were conducted in order to validate the acoustic velocity estimation approach and the uncertainty estimates derived. While in previous studies estimates were obtained using laser Doppler velocimetry (LDV), it is demonstrated that high-repetition rate particle image velocimetry (PIV) can also be successfully employed. The two measurement techniques provide very similar acoustic velocity amplitude and phase estimates for the cases investigated, that are of practical interest for acoustic liner studies. In a broader sense, this approach may be beneficial for non-intrusive sound emission studies in wind tunnel testings.

Integrated cosmological probes: Extended analysis

Recent progress in cosmology has relied on combining different cosmological probes. In earlier work, we implemented an integrated approach to cosmology where the probes are combined into a common framework at the map level. This has the advantage of taking full account of the correlations between the different probes, to provide a stringent test of systematics and of the validity of the cosmological model. We extend this analysis to include not only CMB temperature, galaxy clustering, weak lensing from SDSS but also CMB lensing, weak lensing from the DES SV survey, Type Ia SNe and $H_{0}$ measurements. This yields 12 auto- and cross-power spectra as well as background probes. Furthermore, we extend the treatment of systematic uncertainties. For $\Lambda$CDM, we find results that are consistent with our earlier work. Given our enlarged data set and systematics treatment, this confirms the robustness of our analysis and results. Furthermore, we find that our best-fit cosmological model gives a good fit to the data we consider with no signs of tensions within our analysis. We also find our constraints to be consistent with those found by WMAP9, SPT and ACT and the KiDS weak lensing survey. Comparing with the Planck Collaboration results, we see a broad agreement, but there are indications of a tension from the marginalized constraints in most pairs of cosmological parameters. Since our analysis includes CMB temperature Planck data at $10 < \ell < 610$, the tension appears to arise between the Planck high$-\ell$ and the other measurements. Furthermore, we find the constraints on the probe calibration parameters to be in agreement with expectations, showing that the data sets are mutually consistent. In particular, this yields a confirmation of the amplitude calibration of the weak lensing measurements from SDSS, DES SV and Planck CMB lensing from our integrated analysis. [abridged]

Fractal analysis and statistics of seismic generation rates: the example of the southern California

The seismic generation rate, SGR, at southern California along the 1981–2007 recording period is analysed with the main purpose of finding out whether there exist some correlations between seismic activity before, after and along aftershock sequences triggered by mainshocks of high magnitude. The possibility that a mainshock could be triggered by another neighbouring mainshock and its aftershock sequence is also investigated. The analyses are based on monthly SGR series, obtained as the number of events detected every month along the recording period considered. These monthly SGR series are derived for three aftershock areas associated with Landers (June 28, 1992, Mw = 7.3), Northridge (January 17, 1994, Mw = 6.7) and Hector Mine (October 16, 1999, Mw = 7.1) mainshocks. The most relevant features of SGR series are investigated through various techniques: 1) the rescaled range analysis and the interpretation of the Hurst exponent in terms of persistence, anti-persistence and randomness; 2) time trend estimation by the Kendall-tau algorithm and assessment of their statistical significance by the Mann-Kendall test; 3) the self-affine character, derived from semivariograms, and the Hausdorff measure; 4) autocorrelation and power spectra; 5) cross-correlation and cross-power spectra; 6) the search for the statistical distribution best reproducing the empirical probability of SGR series. Additionally, a close look at plots of epicenters within the aftershock areas, distinguishing between periods of background and aftershock activity, permits detecting some features of the seismicity. Changes on spatial patterns of seismicity suggest that the effects of tectonic stress redistribution could persist beyond an aftershock period, at short and medium distances of the mainshock. This possibility would be also in agreement with cross-correlation results for SGR series.

The cross-correlation between 21 cm intensity mapping maps and the Lyα forest in the post-reionization era

We investigate the cross-correlation signal between 21cm intensity mapping maps and the Lyα forest in the fully non-linear regime using state-of-the-art hydrodynamic simulations. The cross-correlation signal between the Lyα forest and 21cm maps can provide a coherent and comprehensive picture of the neutral hydrogen (HI) content of our Universe in the post-reionization era, probing both its mass content and volume distribution. We compute the auto-power spectra of both fields together with their cross-power spectrum at z = 2.4 and find that on large scales the fields are completely anti-correlated. This anti-correlation arises because regions with high (low) 21cm emission, such as those with a large (low) concentration of damped Lyα systems, will show up as regions with low (high) transmitted flux. We find that on scales smaller than k ≃ 0.2 hMpc−1 the cross-correlation coefficient departs from −1, at a scale where non-linearities show up. We use the anisotropy of the power spectra in redshift-space to determine the values of the bias and of the redshift-space distortion parameters of both fields. We find that the errors on the value of the cosmological and astrophysical parameters could decrease by 30% when adding data from the cross-power spectrum, in a conservative analysis. Our results point out that linear theory is capable of reproducing the shape and amplitude of the cross-power up to rather non-linear scales. Finally, we find that the 21cm-Lyα cross-power spectrum can be detected by combining data from a BOSS-like survey together with 21cm intensity mapping observations by SKA1-MID with a S/N ratio higher than 3 in k∊[0.06,1] hMpc−1. We emphasize that while the shape and amplitude of the 21cm auto-power spectrum can be severely affected by residual foreground contamination, cross-power spectra will be less sensitive to that and therefore can be used to identify systematics in the 21cm maps.

Planck Lensing and Cosmic Infrared Background Cross-correlation with Fermi-LAT: Tracing Dark Matter Signals in the Gamma-Ray Background

Abstract The extragalactic γ-ray background and its spatial anisotropy could potentially contain a signature of dark matter (DM) annihilation or particle decay. Astrophysical foregrounds, such as blazars and star-forming galaxies (SFGs), however, dominate the γ-ray background, precluding an easy detection of the signal associated with the DM annihilation or decay in the background intensity spectrum. The DM imprint on the γ-ray background is expected to be correlated with large-scale structure tracers. In some cases, such a cross-correlation is even expected to have a higher signal-to-noise ratio than the auto-correlation. One reliable tracer of the DM distribution in the large-scale structure is lensing of the cosmic microwave background (CMB), and the cosmic infrared background (CIB) is a reliable tracer of SFGs. We analyze Fermi-LAT data taken over 92 months and study the cross-correlation with Planck CMB lensing, Planck CIB, and Fermi-γ maps. We put upper limits on the DM annihilation cross-section from the cross-power spectra with the γ-ray background anisotropies. The unbiased power spectrum estimation is validated with simulations that include cross-correlated signals. We also provide a set of systematic tests and show that no significant contaminations are found for the measurements presented here. Using γ-ray background map from data gathered over 92 months, we find the best constraint on the DM annihilation with a 1 σ confidence level upper limit of 10−25–10−24 cm3 s−1, when the mass of DM particles is between 20 and 100 GeV.

Intensity Mapping of Hα, Hβ, , and Lines at z &lt; 5

Abstract Intensity mapping is becoming a useful tool to study the large-scale structure of the universe through spatial variations in the integrated emission from galaxies and the intergalactic medium. We study intensity mapping of the , [O iii] 5007 Å, [O ii] 3727 Å, and lines at . The mean intensities of these four emission lines are estimated using the observed luminosity functions (LFs), cosmological simulations, and the star formation rate density (SFRD) derived from observations at . We calculate the intensity power spectra and consider the foreground contamination of other lines at lower redshifts. We use the proposed NASA small explorer SPHEREx (the Spectro-Photometer for the History of the universe, Epoch of Reionization, and Ices Explorer) as a case study for the detectability of the intensity power spectra of the four emission lines. We also investigate the cross-correlation with the 21 cm line probed by the Canadian Hydrogen Intensity Mapping Experiment (CHIME), Tianlai experiment and the Square Kilometer Array (SKA) at . We find both the auto and cross power spectra can be well measured for the Hα, [O iii] and [O ii] lines at , while it is more challenging for the Hβ line. Finally, we estimate the constraint on the SFRD from intensity mapping, and find we can reach an accuracy higher than 7% at , which is better than with the usual method of measurements using the LFs of galaxies.

The improvements of Chang&amp;rsquo;E series satellites orbit determination from GRAIL lunar gravity model

Here we analyze different features between the gravity models of GL0420A and LP165P, including the cross-power spectra of gravity and lunar free-air gravity anomaly. The results show that GL0420A has greatly improved the order and the accuracy of the lunar gravity model, and GL0420A can better reveal the characteristics of the moon. Orbit determination for Chang’E series lunar probe at 100 km×100 km and 100 km×15 km orbit phases is performed with different gravity field models. The results show a better orbit solution with 10 m-level improvement that can be obtained using GL0420A and the solution is more suitable for high accuracy orbit determination for lunar probes.

Cross Spectral Analysis of CODAR-SeaSonde Echoes from Sea Surface and Ionosphere at Taiwan

It is well known that the primary targets responsible for first-order sea echoes observed by a High-Frequency (HF) radar are the advancing and receding ocean waves with the wavelengths at Bragg scales. However, in light of the fact that the ionospheric sporadic E (Es) and F layers may be present in the viewing range of the HF radar for ocean wave detection, the radar returns reflected from the F and Es layers may significantly contaminate the ocean wave power spectrum. The characteristics of the first-order sea echoes and ionospheric interferences measured by the CODAR-SeaSonde in Taiwan area are analyzed and presented in this article. The coherences and phases of the normalized cross spectra of the sea and ionospheric echoes between different pairs of the receiving channels are calculated, respectively. One of the striking features presented in this report is that the ionospheric echo heights scaled from the ionogram observed by the Chung-Li ionosonde are about 30 km lower than those observed by the DATAN CODAR-SeaSonde. It is also found that the coherences of the sea echoes are generally smaller than those of the ionospheric echoes by about 15% on average, and the phase fluctuations (standard deviations) of the sea echoes are substantially larger than those of the ionospheric layer reflection echoes. In addition, statistics show that the sum of the mean phases of the ionospheric echoes between the three receiving channel pairs is approximately zero, while it is not for the sea echoes. These results seem to suggest that the use of the discrepancies in the characteristics of the coherences and phases between the sea and ionospheric echoes may provide a potential means to be helpful to distinguish the sea and ionospheric echoes in the CODAR-SeaSonde observed cross power spectrum.

複数フレームのCSP係数を用いたTDOA推定による複数人物定位·追跡

In this paper, multi-speaker localization methods using voice signal are proposed. Sound source localization systems consisting of many microphones have been widely researched. In our laboratory, we have researched sound source localization of persons based on TDOA (Time Difference of Arrival) between distributed microphones for the purpose of efficient use of air conditioner and lighting in the office . Localization systems based on sound such as footsteps or voice are superior to those based on image data in terms of privacy-consciousness and utility in the dark. As sound source localization methods, beamforming, subspace methods and TDOA-based methods are widely used. CSP (Cross-power Spectrum Phase) analysis is one of the most famous method to estimate TDOA. We have researched sound source localization methods using CSP analysis because it spends less calculation cost than beamforming or subspace methods. However, CSP analysis doesn't work well by itself in the real office environment because of sound lasting for a long time, reverberations, low SNR, and so on. The main cause of difficulty in sound source localization in such a case is the “weak correlation” between different sound signals. In this paper, it is shown that the bad effect of the “weak correlation” can be improved by our proposed methods, Wiener-filter-based CSP analysis and H-CSP (Histogram-CSP) analysis. H-CSP analysis is a statistics data analysis method for CSP coefficients which uses the histogram of CSP coefficients. In this research, the effectiveness of proposed methods are shown by sound source localization experiments in the case sound sources are stable and the case sound sources are moving.

Investigating the hydrodynamics of high temperature fluidized bed by recurrence plot

Effect of temperature on the hydrodynamics of bubbling gas–solid fluidized beds was investigated using recurrence plot (RP) and recurrence quantification analysis (RQA) due to their capabilities for determining the whole bed complexity in a simple way. For this purpose, pressure fluctuations of a fluidized bed of sand particle were measured at various gas velocities and temperatures. Recurrence quantification analysis showed that by increasing the temperature up to 300°C, both determinism and laminarity increase due to formation of larger bubbles whereas entropy and recurrence rate decrease. To better detecting the different structures of in the bed recurrence plot at higher temperature, pressure fluctuations were also decomposed through wavelet transform. It was shown that at low gas velocity, the macro structures became dominate with increasing the bed temperature due to increase in the rate of bubble coalescence. The bubble diameter was estimated from the incoherent component of the cross power spectra of pressure signals at various temperatures and velocities. The incoherent results and standard deviation of measured pressure fluctuations confirmed that by increasing the bed temperature up to 300°C, bubbles grow up to a maximum diameter, after which they became smaller. In addition, the trend of the largest positive Lyapunov exponent was estimated through the recurrence plot patterns. It was shown that a bubbling fluidized bed is a chaotic system at higher temperatures. It was found a maximum in the estimated largest positive Lyapunov exponent at 300°C corresponds to the larger bubbles.

Phase-resolved velocity flow fields induced by laboratory-generated plunging water waves

Digital images of regular plunging water waves were captured by a camera and analyzed to examine the spatial and temporal evolution of phase-resolved instantaneous and ensemble-averaged velocity fields induced by the waves as they propagate and break in the surf zone of a laboratory flume. A computer-driven electronic measurement system was designed, developed and employed to capture images of the breaking waves as they propagate along a glass-walled laboratory flume and break on 1:20 plane slope. An 8-bit monochromatic, progressive scan digital camera was connected to the computer and mounted on the side of the flume to capture images of the breaking waves. Through the use of a trigger pulse from the wave generator, the computer synchronizes image acquisition by the camera at the instance the computer drives the strobe lights to illuminate the field of view. A full wave cycle of the breaking wave was too large to be imaged in one video frame, so the cycle was sub-divided into 20 overlapping phases. For a particular phase, 100 sequential image pairs were captured and saved on a computer. A digital correlation image velocimetry technique was later employed to calculate the spatial cross correlation of the gray scale image data by means of computing the cross-power spectrum of the Fourier transformed image samples. Displacements were measured to subpixel accuracy by interpolating the position of the cross-correlation peak. This gave a combined displacement of polystyrene particles and air bubbles entrained in the flow. One hundred instantaneous velocity flow fields were computed for each phase. Averaging data sets at each phase allowed an estimation of mean velocities at that phase. Structures of the velocity distributions are presented, both as vector fields and in the form of contours plots for phases where turbulence is predominant. Horizontal velocity is up to 2.32c while the vertical is 0.37c, where c=1.08m/s is the wave phase velocity.

DISSECTING THE HIGH-z INTERSTELLAR MEDIUM THROUGH INTENSITY MAPPING CROSS-CORRELATIONS

ABSTRACT We explore the detection, with upcoming spectroscopic surveys, of three-dimensional power spectra of emission line fluctuations produced in different phases of the interstellar medium (ISM) by forbidden transitions of ionized carbon [C ii] (157.7 μm), ionized nitrogen [N ii] (121.9 and 205.2 μm), and neutral oxygen [O i] (145.5 μm) at redshift z &gt; 4. These lines are important coolants of both the neutral and the ionized medium, and probe multiple phases of the ISM. In the framework of the halo model, we compute predictions of the three-dimensional power spectra for two different surveys, showing that they have the required sensitivity to detect cross-power spectra between the [C ii] line and both the [O i] line and the [N ii] lines with sufficient signal-to-noise ratio. The importance of cross-correlating multiple lines with the intensity mapping technique is twofold. On the one hand, we will have multiple probes of the different phases of the ISM, which is key to understanding the interplay between energetic sources, and the gas and dust at high redshift. This kind of study will be useful for a next-generation space observatory such as the NASA Far-IR Surveyor, which will probe the global star formation and the ISM of galaxies from the peak of star formation to the epoch of reionization. On the other hand, emission lines from external galaxies are an important foreground when measuring spectral distortions of the cosmic microwave background spectrum with future space-based experiments like PIXIE; measuring fluctuations in the intensity mapping regime will help constrain the mean amplitude of these lines, and will allow us to better handle this important foreground.