Beam selection system and method

Abstract

The UE determines the power management maximum power reduction P‑MPR for different candidate UE beams. Then, during P2 beam scanning, instead of the UE generating a UE RX beam as wide as possible, the UE considers the P‑MPR determined for the candidate UE beams and generates a UE RX beam with the largest possible beamwidth under the condition of reducing the antenna gain in the UE beam direction with a large P‑MPR.

Description

Technical Field

[0001] Disclosed are embodiments directed to selecting a beam from a set of candidate beams. Background Art

[0002] Narrow beam transmission and reception schemes are typically required at higher frequencies to compensate for high propagation losses. For a given communication link, beams may be applied at a transmit / receive point (TRP) (i.e., an access point, such as a base station, or a component of an access point including an antenna arrangement and a radio unit) and a user equipment (UE) (i.e., a communication device capable of wireless communication with the TRP, such as a smartphone, sensor, home appliance, tablet, etc.). A beam pair consisting of a beam applied at the TRP and a beam applied by the UE is referred to in this disclosure as a "beam pair link (BPL)".

[0003] refer to Figure 1 , the beam management process is used to discover and maintain TRP 104 beam 112 (e.g., TRP transmit (TX) beam) and / or UE 102 beam 116 (e.g., UE receive (RX) beam). Figure 1 In the example of , one BPL has been discovered (i.e., the beam pair consisting of TRP beam 112 and UE beam 116) and maintained by the network. It is expected that the BPL is primarily discovered and monitored by the network using measurements of downlink (DL) reference signals (RS) used for beam management, such as CSI-RS (channel state information RS). CSI-RS for beam management can be sent periodically, semi-persistently, or aperiodically (event triggered), and they can be shared among multiple UEs or specific to a UE. In order to find a suitable TRP TX beam, TRP 104 transmits CSI-RS in different TRP TX beams, and UE 102 performs RSRP measurements on these TRP TX beams and reports back the N best TRP TX beams (where N can be configured by the network). In addition, CSI-RS transmissions on a given TRP TX beam can be repeated to allow the UE to evaluate a suitable UE beam (UE RX beam training).

[0004] There are basically three different implementations of beamforming at the TRP and UE: 1) analog beamforming, 2) digital beamforming, and 3) hybrid beamforming. Each implementation has its advantages and disadvantages. Digital beamforming is the most flexible solution, but it is also the most expensive due to the large number of radios and baseband chains required. Analog beamforming is the least flexible, but is cheaper in terms of manufacturing due to the reduced number of radios and baseband chains. Hybrid beamforming is a compromise between analog beamforming and digital beamforming. At both the TRP and the UE, one type of beamforming antenna architecture that has been agreed to be studied in 3GPP for new radio (NR) access technology is the concept of antenna panels. An antenna panel (or simply "panel") is an antenna array of single-polarized or dual-polarized antenna elements, typically with one transmit / receive unit (TX / RU) per pole. An analog distribution network with phase shifters is used to steer the beam of each panel. Figure 2A An example of a two-dimensional dual polarization panel is shown, and Figure 2B An example of a one-dimensional dual-polarization panel is shown, where each panel is connected to one TXRU on each pole.

[0005] P2 and P3 beam scanning

[0006] It is contemplated that the determination of the BPL between the TRP 104 and the UE 102 (i.e., finding a suitable TRP beam and UE beam) begins with the P2 beam scan, which is performed in Figure 3A . P2 beam scanning is a process that includes the TRP 104 transmitting CSI-RS in different TRP TX directional beams 303 and the UE 102 measuring the RSRP of each directional beam (or simply "beam") in the TRP beam set 303 (using a fixed UE RX beam 301) and sending the (one or more) CSI-RS resource index (CRI) corresponding to the highest (one or more) RSRP back to the TRP 104, where each CRI corresponds to one of the TRP TX beams 303. During such P2 beam scanning, it is expected that, for each UE antenna panel, the UE 102 will apply a wide beam 301 (also called a non-directional beam or omni-directional beam) (e.g., the widest beam that the UE is capable of generating) in order to capture as many propagation paths between the TRP 104 and the UE 102 as possible through the P2 beam scanning. After the P2 beam scanning, the TRP is expected to perform a P3 beam scan, which is performed in Figure 3B. P3 beam scanning is a process that includes TRP 104 sending a CSI-RS burst in (usually) the best reported TRP TX directional beam from the P2 process (which is TRP TX beam 312 in this example), which means that UE 102 can scan through different UE RX directional beams 305, perform measurements on the CSI-RS and select a preferred UE RX directional beam (e.g., UE RX beam 314). How the UE determines the candidate UE RX beams depends on the UE implementation.

[0007] Maximum Permissible Exposure (MPE) for FR2

[0008] Considering the safety of mmWave radiation exposure to the human body, the US Federal Communications Commission (FCC) and other regulatory agencies have defined the maximum permissible exposure as a power density in W / m2. To comply with this, RAN4 has started discussing something called Power Management Maximum Power Reduction (P-MPR), which basically states the amount by which the UE needs to reduce its output power due to RF exposure requirements.

[0009] Since transmissions at mmWave frequencies are expected to be directional, the allowed output power of a UE is likely to be different between different candidate uplink beams across different UE panels. That means that some beams (e.g., beams that may be pointed at a human body) may have potentially very high P-MPRs, while some other beams (e.g., beams whose beam patterns may not coincide with a human body) may have very low P-MPRs. Summary of the invention

[0010] There are certain challenges. During the P2 TRP TX beam scan, the UE should measure the RSRP of different candidate TRP TX beams and report the N TRP TX beams with the highest RSRP. If the beam correspondence is met at both the TRP and the UE, it should be expected that the TRP will configure the UE with UL transmission in the same BPL as the BPL used for DL ​​transmission (i.e., the UE TX beam used for UL transmission is the same as the UE RX beam used for DL ​​transmission). However, this may result in a suboptimal configuration with respect to UL transmission because the configured UE TX beam may have a large P-MPR. In the worst case, due to the potential large P-MPR for the configured UE TX beam, the configured UE TX beam is not allowed to use enough output power to achieve UL coverage, which may cause the UE to lose connection with the network.

[0011] Therefore, the present disclosure proposes that the UE determines the P-MPR for different candidate UE beams. For example, for each candidate beam, the UE determines a value indicating the amount by which the UE should reduce its output power. Then, during the P2 beam scan, instead of the UE generating a UE RX beam that is as wide as possible, the UE considers the P-MPR determined for the candidate UE beam and generates a UE RX beam with as large a beam width as possible under the condition of reducing the antenna gain in the UE beam direction with a large P-MPR. This means that the selected optimal TRP TX beam will not only be selected based on the measured DL RSRP, but will also take into account the UL link budget due to the P-MPR. A significant advantage of this proposal is that the TRP TX beam can be selected while taking into account the UL available output power, which will help the TRP select a BPL that is useful for both UL and DL.

[0012] In one aspect, therefore, a method for beam selection is provided, wherein the method is performed by a UE. In one embodiment, the method includes: the UE determines a first power management level P1 for a first directional beam pointing in a first direction, and determines a second power management level P2 for a second directional beam pointing in a second direction. The method also includes: the UE generates a beam having a first gain in the first direction and a second gain in the second direction using P1 and P2, wherein the first gain is a function of at least P1 and the second gain is a function of at least P2. The method also includes: the UE receives a downlink reference signal sent by a network node using the generated beam. BRIEF DESCRIPTION OF THE DRAWINGS

[0013] The accompanying drawings, which are incorporated herein and form a part of the specification, illustrate various embodiments.

[0014] Figure 1 A wireless communication system is shown.

[0015] Figure 2A An example of a two-dimensional dual polarization panel is shown.

[0016] Figure 2B An example of a one-dimensional dual polarization panel is shown.

[0017] Figure 3A P2 beam scanning is shown.

[0018] Figure 3B P3 beam scanning is shown.

[0019] Figure 4 Problems with conventional beam management procedures are shown.

[0020] Figure 5 A process according to an embodiment is shown.

[0021] Figure 6 is a message flow diagram illustrating a process according to an embodiment.

[0022] Figure 7 is a flow chart illustrating a process according to an embodiment.

[0023] Figure 8 is a block diagram of a UE according to an embodiment.

[0024] Fig. 9 is a module block diagram of a UE according to an embodiment. DETAILED DESCRIPTION

[0025] As described above, if beam correspondence is satisfied at both the TRP and the UE, it is possible that the UE is configured to use a UE TX beam with a large P-MPR for UL transmission. Figure 4 . In step 1, UE 102 determines which UE beams are pointed at the human body and therefore which UE beams experience a large P-MPR. This can be done, for example, by using a camera at the UE, a proximity sensor, etc. It can be seen that UE beam b3 is pointed exactly at the human body and therefore has a large P-MPR. In the next step (step 2), TRP 104 performs a P2 scan so that the UE can report the best N TRP TX beams (e.g., N is greater than or equal to 1). In this case, the UE uses a wide UE RX beam 401 that is as wide as possible in order to properly evaluate all different TRPTX beams. Wide beam 401 is also called an omnidirectional beam because beam 401 can be used to receive signals arriving at the UE from multiple directions. As can be seen in Figure 4 As can be seen in FIG, the best TRP beam is TRP beam b4, which uses a line-of-sight (LOS) path between the TRP and the UE (through the human body). In the next step (step 3), the TRP performs a P3 beam scan to enable UE 104 to find a suitable UE beam. Since only UE beam b3 points in the LOS direction and the other UE beams have very low gain in this direction, the UE must select UE beam b3. However, due to the high P-MPR of this beam, the UE may not be able to transmit at a high enough output power to achieve UL coverage.

[0026] Therefore, the present disclosure proposes an improved UE 502 for determining P-MPR for different candidate UE beams (see Figure 5 ). Then, during P2 beam scanning, instead of the UE generating a wide UE RX beam 401, the UE considers the P-MPR determined for the candidate UE beam and generates a UE RX beam 501 with as large a beam width as possible under the condition of reducing the antenna gain in the UE beam direction with a large P-MPR (see Figure 5). That is, like beam 401, beam 501 is omnidirectional, but for each direction associated with a high P-MPR, the antenna gain is reduced compared to the antenna gain in other directions. This means that the selected best TRP TX beam will not only be selected based on the measured DL RSRP, but will also take into account the UL link budget due to the P-MPR.

[0027] Figure 5 An illustration of the process described above is provided. Before the P2 beam scan begins, the UE 502 determines whether the UE beam is pointing at a human body for each UE beam included in the candidate UE beam set. In this way, the UE 502 is therefore able to determine which UE beams have a large P-MPR. This can be done, for example, by using a camera at the UE, a proximity sensor, etc. It can be seen that the UE beam b3 is pointing just at the human body and therefore has a large P-MPR.

[0028] Then, the TRP 104 serving the UE 502 initiates a P2 beam scan. But now, instead of the UE applying a wide UE RX beam 401, the UE 502 creates a beam 501 that, in one embodiment, is as wide as possible, but has a suppression gain in the direction of the UE beam(s) with a large P-MPR (in this example, UE beam b3). In one embodiment, the UE 502 considers how large the P-MPR is, and the larger the P-MPR, the more the UE 502 suppresses the UE beam gain in that direction.

[0029] A non-limiting example of how UE 502 may generate beam 501 is provided below. For example, as disclosed above, for each UE beam included in a set of N candidate UE beams (N>1), UE 502 determines whether the UE beam is directed toward a human body. The beams in the set of candidate sets may be represented by a set of receiver beam vectors For illustration purposes, assume that UE 502 has determined the P-MPR level P associated with each receiver beam i in the candidate set. i For example, assuming that: i) beam i is directed at a person's head and the person's head is close to the UE, ii) beam j is directed at another person's head but the person's head is not close to the UE, and iii) beam k is not directed at anyone at all, then the situation may be: P i >P j >P k If the N beam vectors in the set form an orthogonal (and optionally rotated) DFT basis and the antenna panel of UE 502 can be described as a uniform linear array (ULA), then the average value of the receiver beam vectors in the set is One way to generate a beam 501 with suppressed gain in the direction of high P-MPR is to then weight the receiver beam vector in the sum by the inverse of the associated P-MPR level, i.e., Among them, W i =1 / P i Another way to generate a beam 501 with suppressed gain in the direction of high P-MPR is to i If it is greater than the threshold, then set W i =0, otherwise set W i =1.

[0030] Typically, according to an embodiment, a candidate set of at least two directional beams is defined by a first set of beam vectors. The generated beams can then be generated from the first candidate set of at least two directional beams by weighting the first set of beam vectors according to the P-MPR of each beam in the candidate set of directional beams. In other words, the more P-MPR a beam experiences, the smaller the weight it should have (i.e., the smaller the magnitude of the weight).

[0031] Using beam 501, UE 502 determines Figure 5 As shown, the best TRP beam in this case is TRP beam b4, which uses a non-LOS path between the TRP and the UE.

[0032] After finding the best TRP beam, the TRP may perform P3 beam scanning to enable the UE 502 to find a suitable UE beam. Since only UE beam b2 points in the direction of the non-LOS path, the UE 502 will select UE beam b2 for subsequent transmissions. Since UE beam b2 is not pointed at the human body, the UE 502 can utilize the maximum allowed output power for the beam.

[0033] Figure 6 is a message flow diagram illustrating the above process. In the first step, UE 502 determines which UE beams are pointed at a human body and therefore which UE beams experience a large P-MPR. UE 502 then determines a UE RX beam with suppression gain in the direction of (one or more) UE502 beams with a large P-MPR. In the next step, the TRP performs a P2 beam scan and the UE502 applies a new UE RX beam when receiving the CSI-RS. UE 502 then reports back the CRI corresponding to the best TRP TX beam measured with the new UE RX beam. The TRP then performs a P3 beam scan and the UE 502 scans through a set of candidate UE RX beams. In one embodiment, the set of candidate UE RX beams does not include any UE beams with a large P-MPR (e.g., beams pointing in the direction of a person), and in this way only beams that do not have too large a P-MPR are evaluated.

[0034] Figure 7 is a flow chart illustrating a process 700 according to an embodiment. The process 700 may start at step s702.

[0035] Step s702 includes the UE 502 determining a first power management level (P1) (eg, an amount by which the output power should be reduced) for a first directional beam (eg, beam b1) pointing in a first direction.

[0036] Step s704 includes UE 502 determining a second power management level (P2) for a second directional beam (eg, beam b3) pointing in a second direction.

[0037] Step s706 includes UE 502 generating a beam (eg, beam 501) having a first gain in a first direction and a second gain in a second direction using P1 and P2, wherein the first gain is a function of at least P1 and the second gain is a function of at least P2.

[0038] Step s708 includes the UE 502 receiving a downlink reference signal sent by a network node (eg, TRP 104) using the generated beam 501.

[0039] In some embodiments, determining P1 includes determining whether the first directional beam is pointing in the direction of the person, and determining P2 includes determining whether the second directional beam is pointing in the direction of the person. Preferably, if it is determined that a) the first directional beam is not pointing in the direction of the person and b) the second directional beam is pointing in the direction of the person, the first gain is higher than the second gain.

[0040] In some embodiments, the first and second directional beams are included in a set of N candidate beams, where N is greater than or equal to two, and the set of candidate beams consists of a corresponding set of receiver beam vectors {f1, f2, ... f N} represents, wherein f1 corresponds to the first directional beam and f2 corresponds to the second directional beam, generating the beam includes calculating W1×f1 and calculating W2×f2, W1 is a function of P1, and W2 is a function of P2. In some embodiments, generating the beam includes: calculating Where, for i = 1 to N, W i YesP i function, and P i is the power management level determined for the i-th beam in a set of N candidate beams. In some embodiments, for i=1 to N, W i =1 / P i , while in other embodiments, for i=1 to N, if P i is less than the threshold, then W i =1, and if Pi If W is greater than the threshold, i =0.

[0041] In some embodiments, receiving a downlink reference signal sent by a network node using the generated beam includes: receiving a reference signal sent by the network node using a first candidate TX directional beam using the generated beam; and receiving a reference signal sent by the network node using a second candidate TX directional beam using the generated beam. In some embodiments, process 700 further includes the following steps: UE 502 obtains a first received reference signal power value PV1 based on a reference signal sent by the network node using the first candidate TX directional beam; UE 502 assigns a rank to the first candidate TX directional beam using PV1; UE 502 determines whether to include a beam indicator (e.g., CRI) indicating the first candidate TX directional beam in a beam report based on the rank assigned to the first candidate TX directional beam; and UE 502 sends a beam report to the network node.

[0042] Figure 8 is a block diagram of a UE 502 according to some embodiments. Figure 8As shown, the UE 502 may include: a processing circuit (PC) 802, which may include one or more processors (P) 855 (e.g., one or more general-purpose microprocessors and / or one or more other processors, such as application-specific integrated circuits (ASICs), field programmable gate arrays (FPGAs), etc.); a communication circuit 848, which is coupled to an antenna arrangement 849 including one or more antennas and includes a transmitter (Tx) 845 and a receiver (Rx) 847 for enabling the UE 502 to send data and receive data (e.g., wireless transmission / reception of data); and a local storage unit (also referred to as a "data storage system") 808, which may include one or more non-volatile storage devices and / or one or more volatile storage devices. In an embodiment where the PC 802 includes a programmable processor, a computer program product (CPP) 841 may be provided. The CPP 841 includes a computer-readable medium (CRM) 842, which stores a computer program (CP) 843 including computer-readable instructions (CRI) 844. CRM842 can be a non-transitory computer-readable medium, such as a magnetic medium (e.g., a hard disk), an optical medium, a memory device (e.g., a random access memory, a flash memory), etc. In some embodiments, the CRI 844 of the computer program 843 is configured so that when executed by the PC 802, the CRI causes the UE 502 to perform the steps described herein (e.g., the steps described in the reference flow chart herein). In other embodiments, the UE 502 can be configured to perform the steps described herein without the need for code. That is, for example, the PC 802 can consist of only one or more ASICs. Therefore, the features of the embodiments described herein can be implemented in hardware and / or software.

[0043] Fig. 9 5 is a schematic block diagram of a UE 502 according to some other embodiments. In some embodiments, the UE 502 includes one or more modules, each of which is implemented in software. The module(s) provide the functionality described herein (e.g., Figure 7In one embodiment, the module includes: a determination module 902, which is configured to: determine a first power management level P1 for a first directional beam (e.g., b1, b2, or b4) pointing in a first direction, and determine a second power management level P2 for a second directional beam (e.g., b3) pointing in a second direction; a beam forming module 904, which is configured to: use P1 and P2 to generate a beam (e.g., beam 501) having a first gain in the first direction and a second gain in the second direction, wherein the first gain is a function of at least P1 and the second gain is a function of at least P2; and a receiver module 906, which is configured to use the generated beam to receive a downlink reference signal sent by a network node (e.g., TRP 104).

[0044] Although various embodiments of the present disclosure are described herein, it should be understood that they are presented in an exemplary and non-restrictive manner only. Therefore, the width and scope of the present disclosure should not be limited by any exemplary embodiment described above. Generally, all terms used in this article will be interpreted according to their common meanings in the relevant technical field, unless different meanings are clearly given and / or different meanings are implied in the context of using it. Unless otherwise clearly stated, all references to elements, devices, components, methods, steps, etc. will be openly interpreted as referring to at least one example in elements, devices, components, methods, steps, etc. Unless otherwise specified herein or obviously contradictory with the context, any combination of the above-described elements in all possible variations thereof is covered by the present disclosure.

[0045] Furthermore, while the processes described above and shown in the accompanying drawings are shown as a series of steps, this is done for illustration purposes only. Therefore, it is contemplated that some steps may be added, some steps may be omitted, the order of the steps may be rearranged, and some steps may be performed in parallel. That is, the steps of any method disclosed herein do not necessarily have to be performed in the exact order disclosed, unless a step is explicitly described as being after or before another step and / or it is implied that a step must be after or before another step.

Claims

1. A method for beam selection, the method being performed by a user equipment UE, the method comprising: For each candidate beam vector (f l ,f2,...,f N ), wherein N is not less than 4, determining a power management maximum power reduction P-MPR level for the candidate beam vector, thereby determining N P-MPR levels, one P-MPR level for each candidate beam vector, wherein each candidate beam vector generates a beam pointing in a different direction, such that a first candidate beam vector generates a first candidate beam pointing in a first direction, a second candidate beam vector generates a second candidate beam pointing in a second direction, a third candidate beam vector generates a third candidate beam pointing in a third direction, a fourth candidate beam vector generates a fourth candidate beam pointing in a fourth direction, and each P-MPR level is associated with one of the different directions; For each candidate beam vector, a weight (W ) for the candidate beam vector is determined based on the P-MPR level determined for the candidate beam vector. l ,W2,...,W N ), thereby determining N weights, one weight for each candidate beam vector; generating an omnidirectional beam using the N candidate beam vectors and the N weights such that the generated omnidirectional beam has a suppressed gain in a direction associated with a candidate beam vector having a P-MPR level exceeding a threshold; and receiving a downlink reference signal sent by a network node using the generated omni-directional beam.

2. The method according to claim 1, wherein: Generating the omnidirectional beam comprises: calculate Where, for i = 1 to N, W i YesP i function, and P i is the P-MPR level determined for the i-th candidate beam vector.

3. The method according to claim 2, wherein: For i = 1 to N, W i =1 / P i .

4. The method according to claim 2, wherein: For i = 1 to N, if P i is less than the threshold, then W i =1, and if P i is greater than the threshold, then W i =0.

5. The method according to any one of claims 1 to 4, wherein: Receiving the downlink reference signal sent by the network node using the generated omni-directional beam includes: receiving, using the generated omni-directional beam, a reference signal sent by the network node using a first candidate transmit TX directional beam; and A reference signal sent by the network node using a second candidate transmit TX directional beam is received using the generated omni-directional beam.

6. The method according to claim 5, further comprising: Obtaining a first received reference signal power value PV1 based on a reference signal sent by the network node using the first candidate transmit TX directional beam; Assigning a rank to the first candidate transmit TX directional beam using PV1; determining whether to include a beam indicator indicating the first candidate transmit TX directional beam in a beam report based on the rank assigned to the first candidate transmit TX directional beam; as well as Sending the beam report to the network node.

7. A computer program product comprising instructions which, when executed by a processing circuit of a user equipment, cause the user equipment to perform the method according to any one of claims 1 to 6.

8. A computer-readable storage medium comprising instructions, which, when executed by a processing circuit of a user equipment, cause the user equipment to perform the method according to any one of claims 1 to 6.

9. A user equipment, the user equipment being adapted to: For each candidate beam vector (f l ,f2,...,f N ), where N is not less than 4, determining a power management maximum power reduction P-MPR level for the candidate beam vector, thereby determining N P-MPR levels, one P-MPR level for each candidate beam vector, where each candidate beam vector generates a beam pointing in a different direction, such that a first candidate beam vector generates a first candidate beam pointing in a first direction, a second candidate beam vector generates a second candidate beam pointing in a second direction, a third candidate beam vector generates a third candidate beam pointing in a third direction, a fourth candidate beam vector generates a fourth candidate beam pointing in a fourth direction, and each P-MPR level is associated with one of the different directions; For each candidate beam vector, a weight (W ) for the candidate beam vector is determined based on the P-MPR level determined for the candidate beam vector. l ,W2,...,W N ), thereby determining N weights, one weight for each candidate beam vector; generating an omnidirectional beam using the N candidate beam vectors and the N weights such that the generated omnidirectional beam has a suppressed gain in a direction associated with a candidate beam vector having a P-MPR level exceeding a threshold; as well as A downlink reference signal transmitted by the network node is received using the generated omni-directional beam.

10. The user equipment according to claim 9, wherein: The user equipment is adapted to generate the omnidirectional beam by performing a process comprising: calculate Where, for i = 1 to N, W i YesP i function, and P i is the P-MPR level determined for the i-th candidate beam vector.

11. The user equipment according to claim 10, wherein: For i = 1 to N, W i =1 / P i .

12. The user equipment according to claim 10, wherein: For i = 1 to N, if P i is less than the threshold, then W i =1, and if P i is greater than the threshold, then W i =0.

13. The user equipment according to any one of claims 9 to 12, wherein: The user equipment is adapted to receive the downlink reference signal sent by the network node using the generated omni-directional beam by performing a process comprising: receiving, using the generated omni-directional beam, a reference signal sent by the network node using a first candidate transmit TX directional beam; and A reference signal sent by the network node using a second candidate transmit TX directional beam is received using the generated omni-directional beam.

14. The user equipment according to claim 13, wherein: The user equipment is further adapted to: Obtaining a first received reference signal power value PV1 based on a reference signal sent by the network node using the first candidate transmit TX directional beam; Assigning a rank to the first candidate transmit TX directional beam using PV1; determining whether to include a beam indicator indicating the first candidate transmit TX directional beam in a beam report based on the rank assigned to the first candidate transmit TX directional beam; as well as Sending the beam report to the network node.

Images