[NR][UE][RF] Transmitter Characteristics [NR5G 终端射频(发射机)指标,最大发射功率,最大功率回退值,额外最大功率回退值]

本文详细介绍了无线通信中UE(用户设备)的最大发射功率及其在不同功率类别下的最大功率减少(MPR)计算。内容涵盖了数学运算、资源块分类、边沿和内部资源块定义,并针对不同信道带宽给出了具体的MPR值。此外,还讨论了额外最大输出功率减少(AMPR)的情况,以及与网络信号相关的参数。内容基于3GPPTS38.101-1V17.4.0标准。

摘要生成于 C知道 ,由 DeepSeek-R1 满血版支持, 前往体验 >

Transmitter Characteristics

1. Transmitter power

1.1 UE maximum output power

在这里插入图片描述
在这里插入图片描述

1.2 UE maximum output power reduction (MPR)

1. General

Before talking about specifc MPR values in different PCs, let us classify RB types based on RB location.
(1) Firstly, some mathematics operations,

  • max() is the largest value of all arguments
  • floor(x) is the greatest integer less than or equal to x.
  • ceil(x) is the smallest integer greater than or equal to x.

(2) Secondly, some RB locations,

  • NRB is the maximum number of RBs for a given Channel bandwidth and sub-carrier spacing.
  • LCRB is transmission bandwidth which represents the length of a contiguous resource block allocation expressed in units of resources blocks
  • RBStart,Low = max(1, floor(LCRB/2))
  • RBStart,High = NRB – RBStart,Low – LCRB

(3) Lastly, Inner/Edge/Outer RB definition,

  • Inner RB allocation
    RBStart,Low ≤ RBStart ≤ RBStart,High, and
    LCRB ≤ ceil(NRB/2)

  • Edge RB allocation
    An Edge RB allocation is the one for which the RB(s) is (are) allocated at the lowermost or uppermost edge of the channel with
    LCRB ≤ 4 RBs for power class 1.5 and
    LCRB ≤ 2 RBs for other power classes.

  • Outer RB allocation
    The RB allocation is an Outer RB allocation for all other allocations which are not an Inner RB allocation or Edge RB allocation.

2. MPR values in different PCs

For channel bandwidths ≤ 100 MHz,

  • MPR of power class 3 — Table 6.2.2-1

  • MPR of power class 2 — Table 6.2.2-2

  • ∆MPR

    • ∆MPR — Table 6.2.2-3
      if the relative channel bandwidth > 4% for TDD bands or > 3% for FDD bands

    • ∆MPR = 0
      if the relative channel bandwidth ≤ 4% for TDD bands or ≤ 3% for FDD band, the ∆MPR is set to zero

      Where relative channel bandwidth = 2*BWChannel / (FUL_low + FUL_high)

  • MPR of power class 1.5— Table 6.2.2-4 and Table 6.2.2-4a

  • MPR of power class 1 in n14 — Table 6.2.2-5

在这里插入图片描述
在这里插入图片描述
在这里插入图片描述
在这里插入图片描述

在这里插入图片描述

1.3 UE additional maximum output power reduction (AMPR)

Each additional emission requirement is associated with a unique network signalling (NS) value indicated in RRC signalling by an NR frequency band number of the applicable operating band and an associated value in the field additionalSpectrumEmission.

the total reduction to UE maximum output power is max(MPR, A-MPR)

1. A-MPR for NS_04

For NS_04, A-MPR is not added to MPR. Also, when NS_04 is signalled, MPR shall be set to zero in the PCMAX equations to avoid double counting MPR. Allowed maximum power reduction is defined as A-MPR = max(MPR, A-MPR’), Note that A-MPR’ = 0 dB means only MPR is applied, where A-MPR’ is defined as

  • if
    RBstart ≤ fstart,max,IMD3 / (12SCS) and
    LCRB ≤ AWmax,IMD3 / (12SCS) and
    FC - BWChannel/2 < FUL_low + offsetIMD3,
    then
    the A-MPR’ is defined according to Table 6.2.3.2-2
    PC3_A2 relative to 23 dBm for power class 3,
    PC2_A4 relative to 26 dBm for power class 2,
    and PC1.5_A6 relative to 29 dBm for power class 1.5,
  • else,if
    RBstart ≤ LCRB/2 + deltastart / (12SCS) and
    LCRB ≤ AWmax,regrowth / (12SCS) and
    FC - BWChannel/2 < FUL_low + offsetregrowth,
    then
    the A-MPR’ is defined according to Table 6.2.3.2-2
    PC3_A1 relative to 23 dBm for power class 3,
    PC2_A3 relative to 26 dBm for power class 2,
    and PC1.5_A5 relative to 29 dBm for power class 1.5,
  • else
    A-MPR’ = 0 dB and apply MPR.

With the parameters defined in Table 6.2.3.2-1.
在这里插入图片描述
在这里插入图片描述

1.4 Configured transmitted power

The UE is allowed to set its configured maximum output power PCMAX,f,c for carrier f of serving cell c in each slot. The configured maximum output power PCMAX,f,c is set within the following bounds:
PCMAX_L,f,c ≤ PCMAX,f,c ≤ PCMAX_H,f,c with
在这里插入图片描述

Reference

[1] 3GPP TS 38.101-1 V17.4.0 (2021-12)

### Inner-MPDIoU 损失函数概述 Inner-MPDIoU损失函数专门用于处理目标检测中的多部件对象,其计算过程与MPDIoU相似,但仅关注于目标内部各部件间的交并比(IoU)[^1]。 这种设计使得Inner-MPDIoU更适合复杂结构的目标识别任务,在这些场景下,单个物体可能由多个部分组成。通过专注于内部组件而非整体轮廓,可以更精确地捕捉到不同部位之间的相对位置关系,进而提高模型对于细粒度特征的学习能力。 ### Inner-MPDIoU 损失函数实现 以下是Python代码示例,展示了如何基于PyTorch框架实现Inner-MPDIoU损失函数: ```python import torch def inner_mpdiou_loss(pred_boxes, target_boxes, parts_mask=None): """ 计算Inner MPDIoU损失 参数: pred_boxes (Tensor): 预测边界框 [N,M,4] target_boxes (Tensor): 目标边界框 [N,M,4] parts_mask (Tensor or None): 组件掩码 [N,M] 如果为None,则默认所有都是有效部件 返回: Tensor: 平均内含MPDIoU损失 """ def box_area(box): return (box[:, :, 2] - box[:, :, 0]) * (box[:, :, 3] - box[:, :, 1]) def box_iou(boxes1, boxes2): area1 = box_area(boxes1) area2 = box_area(boxes2) lt = torch.max(boxes1[:, :, :2], boxes2[:, :, :2]) rb = torch.min(boxes1[:, :, 2:], boxes2[:, :, 2:]) wh = (rb - lt).clamp(min=0) inter = wh[:, :, 0] * wh[:, :, 1] union = area1 + area2 - inter iou = inter / union return iou N, M, _ = pred_boxes.shape if parts_mask is None: parts_mask = torch.ones((N, M), device=pred_boxes.device) # 计算IoUs矩阵 ious = box_iou(pred_boxes, target_boxes) # 应用mask过滤掉不需要比较的部分 masked_ious = ious * parts_mask.unsqueeze(-1) # 对每张图片求平均IOU作为最终loss loss_per_image = masked_ious.sum(dim=[1, 2]) / parts_mask.sum(dim=-1) return 1 - loss_per_image.mean() ``` 此版本的`inner_mpdiou_loss`接受预测边界框(`pred_boxes`)和真实标签边界框(`target_boxes`)以及可选参数`parts_mask`来指定哪些区域应该参与计算。当未提供`parts_mask`时,默认认为所有的部件都应被考虑进去。 ### Inner-MPDIoU 损失函数应用场景 在实际应用中,Inner-MPDIoU特别适用于那些需要精细定位的对象类别,比如人体姿态估计或多视角三维重建等领域。由于这类任务往往涉及到复杂的几何形状及其组成部分之间微妙的空间布局,因此采用更加细致入微的方法来进行误差衡量显得尤为重要。 此外,考虑到边界框回归过程中IoU的变化情况有助于理解模型性能随时间推移而发生怎样的改变,并据此调整训练策略以达到更好的效果[^2][^3]。
评论
添加红包

请填写红包祝福语或标题

红包个数最小为10个

红包金额最低5元

当前余额3.43前往充值 >
需支付:10.00
成就一亿技术人!
领取后你会自动成为博主和红包主的粉丝 规则
hope_wisdom
发出的红包
实付
使用余额支付
点击重新获取
扫码支付
钱包余额 0

抵扣说明:

1.余额是钱包充值的虚拟货币,按照1:1的比例进行支付金额的抵扣。
2.余额无法直接购买下载,可以购买VIP、付费专栏及课程。

余额充值